If you haven’t checked out the rewards section of the Nvidia app in a while, you might want to do that, especially if you’ve been wanting to try the whole Adobe tool stack without spending a penny. As part of Nvidia’s collaboration with Adobe, the company is giving specific RTX GPU owners access to Adobe Creative Cloud for free, which includes popular photo and video editing tools like Photoshop, Premiere Pro, After Effects, and more.

The offer is currently live in the Nvidia App Rewards section, and to be eligible, you must have an RTX series GPU from any of the last three generations. However, there’s a difference: with a GeForce RTX 30 or 40 series card, you’ll get one free month of Adobe Creative Cloud. Meanwhile, if you own an RTX 50 GPU, Nvidia will treat you to two free months instead.

On top of that, one benefit exclusive to RTX 50 series GPUs is the Substance 3D reward, which gives you access to five different Adobe apps along with a massive asset library for your games, including tools like 3D Sampler, 3D Designer, and 3D Painter. It’s a fairly solid deal overall, even if you consider Adobe Creative Cloud alone, which normally costs $69 per month.

Don’t forget to cancel on time

The classic Tetris game has been packaged into a PDF and dubbed Pdftris. The classic falling blocks puzzler is claimed to be playable in most desktop browsers and, indeed, we successfully loaded and played the game in Firefox on PC simply by clicking the above link. A security analyst named Thomas Rinsma is the brains behind this new PDF-based browser game.

Background details on the project aren't extensive, as the GitHub project page ReadMe entry is empty. However, Rinsma was happy to chat with users on the Hacker News feed about the development project.

Rinsma indicates that he was inspired to port a game to run in a PDF after he "realized that the PDF engines of modern desktop browsers (PDFium and PDF.js) support JavaScript with enough I/O primitives to make a basic game like Tetris."

The security analyst and self-confessed aficionado of "strange loops and weird machines" also explained that it wasn't that easy to get his Tetris game to run correctly in both engines. However, he discovered that "showing/hiding annotation 'fields' works well to make monochrome pixels." Moreover, if you try the game with the on-screen buttons and would prefer to use keys, then Rinsma made it so that keyboard control can be achieved by typing in a text input box below the game area.

Rinsma humbly admits that his Tetris in PDF code is "quite janky." However, he suggests it at least serves as a reminder of the power and flexibility of PDF scripting.

Those interested in prying inside the Pdftris code don't need specialist or esoteric development tools. Just choose to download the 60KB PDF, and open it in a text editor and you will find the file is "all ASCII."

There is some source code available, too, with greater readability and comments inserted. You can see that on the aforementioned GitHub project page, specifically here.

Rinsma noted that Pdftris isn't working on Adobe's Reader "likely because it's not compliant in several ways." Other non-browser Pdftris possibilities like Foxit Reader might work, he mused, as he believes it supports the necessary scripting for the gaming action.

It would be great to get some color, shading, and sound in a future release of Pdftris but we don't know if this project will be developed much further. In some ways, it is basically a proof of concept.

While Pdftris is certainly an eyebrow-raising mashup of the gaming and document rendering worlds, developers have dragged far duller apps into the gaming realm, particularly Microsoft Excel. Microsoft's spreadsheet can even host a 3D raytracing gaming engine (with the help of VBA).

Logitech is getting into the macro pad game with its new MX Creative Console, facing off against established players like Elgato, Razer, and Cooler Master. The Swiss firm's two-piece device is marketed towards Adobe Creative Cloud users, especially those who spend hours on Photoshop, Premiere Pro, or Lightroom Classic daily. According to Logitech MX General Manager Anatoliy Polyanker (via Business Wire), “The MX Creative Console is designed to help people redefine their workflow, enabling them to work smarter and faster.”

The MX Creative Console has two primary parts — a wired keypad with nine primary LED display keys and two arrow keys for navigating between pages, plus a dial pad that has a massive knob, a jog wheel, and four extra buttons. You can pretty much program all the buttons to suit your particular needs, but the company has worked closely with Adobe so that it will work straight out of the box with apps like Lightroom Classic, Photoshop, and Premiere Pro.

At $200, the MX Creative Console is quite an expensive device, especially as you can get a Stream Deck starting at just $60. There are also other options in that price range, like the Razer Stream Controller or the Elgato Stream Deck Plus. However, Logitech is the only one of these big-name firms to offer the dial pad and keypad combination, which makes it highly suitable for creatives working with Adobe apps. The other competitors are more focused on live streaming and audio production, so you’ll have to spend a lot of time configuring them to make them suitable for visual creative workflows.

The Cooler Master MasterHUB is the nearest competitor to the MX Creative Console, but the complete modular kit costs double at $399 and it’s still in the Kickstarter stage at the time of writing (although we expect shipments to begin this month). Furthermore, since it’s not explicitly made for Adobe app workflows, you may have to spend a good amount of time discovering how to program the shortcuts for each app.

While this MX Creative Console is perfect for Adobe users, you might find it lacking if you prefer other apps. You’ll have to customize the console yourself if you want to use it with non-Adobe programs, like DaVinci Resolve or GIMP. Alternatively, you could download plugins via the Plugin Marketplace in Logitech Options+ app to make the MX Creative Console work more seamlessly with select apps like Spotify, Capture One, and Zoom. However, you’ll have to wait for Logitech to see if and when it will release plugins for popular Adobe alternatives.

Adobe PDF files are one of the leading file types for documents around. These proprietary file types work mainly with Adobe software but can be easily converted to other file types, as well, like Word Docs or JPGs. And sometimes you’ll find yourself needing to edit a PDF. If you’ve already got access to Adobe Acrobat, this is easy to do. However, those without Acrobat may be wondering how they can edit a PDF. Thankfully, Adobe provides a free tool that anyone can use to edit PDFs online for free. The only caveat is that you will need to make an account with Adobe to use it.

 How to Edit a PDF in Acrobat 

If you’ve got a copy of Adobe Acrobat, editing a PDF is fairly easy. But if you’re not sure how to enable editing mode, here’s what you have to do.

1.  Open the PDF you want to edit in Adobe Acrobat.

2. Go to the Tools tab in the upper left-hand corner. 

3. Click `Edit PDF`.

4. Edit the PDF using the tools provided.

Don't forget to save your changes when you are done.

How to Edit a PDF for Free

If you don’t have Adobe Acrobat, it’s still possible to edit PDFs. Adobe provides a free online tool that has plenty of features to enable basic changes to most PDF documents. You can change the order of pages, add or remove elements, and even digitally sign documents.

1. Go to https://acrobat.adobe.com/link/acrobat/add-comment. 

2. Create an Adobe account if you don’t already have one. Even though you’re prompted to create an account, it’s free to use the online editing tool.

3. Drag and drop the PDF into the window.

4. Click ‘Edit’ in the upper left corner.

5. Edit the PDF using the tools provided.

These steps should work for anyone whether they’ve got access to Adobe Acrobat or not. The online tool is free for anyone to use but an account with Adobe is required to gain access to it. 

TIP: Even if you’ve got a copy of Acrobat, it’s handy to keep the free online editing tool on hand in case you’re at a machine that doesn’t have access to the full Acrobat suite.

4. Edit the PDF using the tools provided. Save when done.

Adobe’s PDF file is one of the most commonly used file types for documents but what if you need it converted into an image? Thankfully, you’ve got a few options available when it comes to converting a PDF to a JPG or converting a JPG to a PDF. It helps to have Adobe Acrobat but you don’t need it at all to convert files as there are free tools available that convert these files both ways.

How to Convert a PDF File to a JPG

This method requires a copy of Adobe Acrobat. If you do not have Acrobat, you can use the free steps below for an browser-based solution.

1. Open the PDF you want to convert in Adobe Acrobat. 

2. Click File in the upper left corner and go to ‘Export To’ then ‘Image’ and choose ‘JPEG’. 

3. Click ‘Save’ once you choose a location for your new JPG file. 

4. Open the file in an image viewer to make sure it works. 

 How to Convert a PDF file to a JPG for Free 

If you don’t have Acrobat, you can use Adobe’s free online conversion tool to make a JPG from your PDF for free.

1. Go to https://www.adobe.com/acrobat/online/pdf-to-jpg.html. 

2. Drag and drop the PDF onto the browser window 

3. Choose the JPG quality for your new file using the drop down menu. 

4. Click ‘Convert to JPG’ in the bottom right 

5. Click ‘Download File’. The JPG file will save as a .zip file to your PC. 

6. Extract the JPG from the ZIP file. 

If you’re unfamiliar with using ZIP files, check out our guide on how to add and remove content from ZIP files in Windows.

 How to Convert a JPG to a PDF File 

You can convert JPG images into PDF files easily using Acrobat but if you don’t have a copy, you can always use the free steps below to do so using Adobe’s online converter tool.

1. Open the JPG you want to convert in Adobe Acrobat. By default, it will be converted to a PDF file. 

2. Click ‘Save’ after you choose a location for the new file.

 How to Convert a JPG to a PDF File for Free 

Adobe provides a free browser-based tool that will convert JPGs to PDFs for free. This is a great option for those who don’t have Adobe Acrobat.

1. Go to https://acrobat.adobe.com/link/acrobat/jpg-to-pdf.

2. Drag and drop the JPG file you want to convert into the browser window.

3. Click ‘Download’ in the upper right corner and choose a location to save the file.

Adobe makes the conversion process very easy using the integrated conversion tools in Acrobat as well as the online converters. So whether you own Acrobat or not, you can still convert their PDF as needed to the most common file types.

When it comes to macro keyboards, Elgato’s Stream Decks usually come to mind. For many streamers, the highly-customizable input devices, complete with full-color screens under each key, are critical tools that allow them to perform complex actions in OBS  in just one step.

Loupedeck’s line of macro devices take the user well beyond streaming. Don’t get me wrong: Any macro keyboard, even a home-made Raspberry Pi one, can be used for other purposes than streaming — but how many have features integrated for video, photo, or even animation editing? Loupedeck’s consoles are built to work with Twitch, Adobe Suite, Ableton, Twitter, OBS, and music software such as Spotify right out of the box. 

I took a look at the three consoles in Loupedeck’s current lineup: the Loupedeck Live ($269), the Loupedeck+ ($279), and the Loupedeck CT ($549). Loupedeck also recently announced the upcoming Loupedeck S ($180), which is currently fully-funded on Indiegogo and is set to start shipping in the fall. Razer also recently announced its own branded version of the Loupedeck Live which is identical to the original but has Razer branding. 

Loupedeck Live

The Loupedeck Live ($269) is the smallest of Loupedeck’s current lineup and is geared toward streamers and content creators. It has 12 programmable touchscreen buttons, six tactile knobs, and eight tactile buttons (one of which is the dedicated home/back button). I don't know what it is about this black matte finish but dust LOVES it. Luckily there is a lock mode that turns off the touch screen and button functions so you can wipe off the console.

Just like the Loupedeck Live has a lock mode there is something called a Dynamic Mode where the buttons on the touchscreen will switch depending on the application running. I can have Photoshop and Premiere open at the same time, which is kind of overkill but sometimes necessary. My touchscreen buttons will change from photo editing to video editing just by clicking on Premiere. 

The Live has integrations with Twitch that can start a commercial break, welcome a new subscriber with a special sound effect, or simply stop and start your stream. Lined with three knobs on each side and eight round buttons at the bottom, the Live is suited to control volume, skip songs, and scrub through a video. 

It’s great for keyboard shortcuts in any program. For example, buttons can be created for copy and paste or activating full screen. The knobs on the side are called Rotating Dials. They have two functions, rotating infinitely with built-in notches and pressing in. This creates 12 functions out of these six knobs. The eight round buttons at the bottom are called the Workspace buttons, with the first one being the Home or Back button. These buttons help navigate between Workspaces or pages - which are a list of buttons that show on the face of the Loupedeck. There are two ways to move between pages; swiping or by pressing the round buttons to move on to the next page. 

This is the smallest of the Loupedeck consoles but still packs a lot of configuration power. With over 70 native programs supported in software, the  Live is great for any level of streamer, producer, programmer, or computer user — novice to intermediate. This is a piece of hardware that can grow with you if you’re just starting out. As an advanced streamer, I enjoy using the Loupedeck Live to try out new commands, engage viewers, and help streamline the process of streaming. It doesn’t take up too much room on your desk, measuring 5.9 inches wide by 4.3 inches deep, and 1.18 inches thick. It weighs about half a pound. 

The Loupedeck Live really helped me to get organized and categorize my movements and the processes I use to run one of my game streams. I can create one page, or a set of buttons,  for the beginning of my stream, swipe to the next page for the middle of my stream, and then go to a third page to wrap up my stream. And I can create duplicate buttons so I don’t have to switch back and forth between pages.

Once a live stream is over I can move on to post-stream activities, such as opening Adobe Premiere to start working on editing the footage. With the help of the Loupedeck Live’s Dynamic Mode, my pages or workspace change when I open up different applications. If I’d rather wind down instead of getting right to work, I can set up a page for that. I can create different profiles and call them things such as “Stream Junae Profile,” “Post-Stream Junae Profile,” or “Leisure Junae Profile.” And have everything I need within those profiles such as Spotify controls and buttons to launch my favorite games or open productivity apps such as Microsoft Word to write articles — it’s very intuitive.

Loupedeck CT

The $549 Loupedeck CT is made for creative professionals who do serious video or audio editing. The Loupedeck CT is the largest of the Loupedeck consoles (as opposed to the Loupedeck+, which is more of a deck than a console), measuring 6.2 inches wide by 5.9 inches deep and 1.18 inches thick. The Loupedeck CT has the same buttons and knobs as the Loupedeck Live, and adds a bottom section which features a large dial and 12 square tactile buttons.

The Loupedeck CT is bulkier than the Live, tipping the scales at 0.8 pounds, but I wouldn’t change a thing. The black matte finish attracts quite a bit of dust, but is also very easy to clean. The separation of the touchscreen buttons and tactile buttons so I don’t click a function I didn’t intend to use is a great design.

The dial is definitely an upgrade. Every button on the Loupedeck CT has some kind of feedback — haptic or tactile — except for the dial. Baby, when I tell you turning this dial is like spreading whipped butter on a piece of bread with sugar and cinnamon, believe it! It’s so smooth! The dial makes scrolling through web pages, scrubbing through editing timelines, selecting colors, and adjusting highlights and shadows a very smooth, enjoyable process.

The dial and buttons — like everything on Loupedeck’s consoles — are customizable. Adding these buttons relieves the need to create a shortcut function. The 12 buttons include inputs from a keypad, such as up/down, Fn, and Ctrl — no more pressing Ctrl +Z like a caveman! The buttons also include shortcuts such as Undo and Save.

Every button, knob, and dial on the Loupedeck CT is customizable and works fabulously in Dynamic Mode. Dynamic Mode allows you to switch seamlessly between programs and continue to use the Loupedeck CT without interrupting your workflow. For example, I can be using OBS and switch to Ableton, and the buttons and functions on the Loupedeck CT will change with me. If there are multiple people using the Loupedeck CT, you can create individual profiles for each person — or you can create profiles for different work and lifestyle modes.

Some macro keypads limit you to three pages of buttons, but the Loupedeck CT lets you create up to 10 pages of buttons (12 buttons per page). That’s over 100 button functions available on one profile.

It doesn’t matter how many (or few) buttons you have, the Loupedeck CT’s dial is definitely its stand-out feature. I use the dial in photo and video editing to get precise measurements that are easily overshot on a mouse, or that take too long on a keyboard. For example, if I want to turn up an effect to an exact number — such as turning the vignette to 62 — it’s difficult on a mouse, but I can use the dial to get there without having to make precise mouse movements.

The Loupedeck CT does everything the Loupedeck Live does — it’s easy to program it to work with streaming platforms such as OBS and Twitch, and to get it to play SFX for streaming without having to program them in StreamElements. I especially like that the Fn buttons can be placed on the keyboard, as this means less programming.

The Loupedeck CT lets me do all the things I need to do during a stream, and easily integrates with Twitch, Streamlabs, and OBS. But it’s in post-stream content editing where the Loupedeck CT’s jog dial really stands out. Because the Loupedeck CT is larger, heavier, and has a dial — which is potentially easier to damage — I wouldn’t feel comfortable traveling with this bulkier console. But if you can afford it the Loupedeck CT’s versatility makes it a powerful streaming companion.

Loupedeck+

If you want the feel of a professional editing deck like the ones they use in studios, get a Loupedeck+. With more than 40 buttons, knobs, and switches, the Loupedeck+ is largest in the series, measuring 17.8 inches wide, 8.2 inches deep, and 1.9 inches thick, and weighing about 1.5 pounds — it’s roughly the size of a typical PC keyboard.

Unlike the Loupedeck Live and Loupedeck CT consoles, the Loupedeck+ doesn’t have a screen. Instead, it’s got buttons on buttons on buttons, mixed with knobs, dials, and switches — this is an editing machine that editors and producers will love! The Loupedeck+ is advanced enough to do all of the intricate things you need to do in photo, video, and audio editing, and it’s simple enough for a beginner to use.

The press in functions on dials reset their values back to zero — useful if you’re trying to take something such as color saturation back to its default. The color switches make it easier to adjust the greens, blues, and reds for photo and video color correction. The dials make it easier to accurately control the levels of functions rather than trying to guess by looking at sliding bars.

Thanks to notches in each knob, values don’t increase too quickly. Although the buttons on the Loupedeck+ have printed labels such as Temperature, Tint, and Saturation, they’re still customizable — you can overwrite the default functions to perform whatever functions you like, though you’ll have to remember what you changed them to (or, I suppose, you could re-label them).

The Loupedeck+ has built-in level buttons that you can use for different levels of editing. For example, Level 1 could be the default settings button, Level 2 could be all color gradients, and Level 3 could be for adding extra effects. Again, each button is still completely customizable.

If you’re new to editing photos and videos, the Loupedeck+ might be overwhelming at first. I’m not new to editing, but I don’t think the Loupedeck+ is an ideal console for me because while I like to edit, I also like to stream and have shortcuts on my console that open up regular applications such as email. The Loupedeck+ is highly programmable, just like the other Loupedeck consoles, but it’s definitely designed for editing photos and videos. It’s the cheapest of the three Loupedeck consoles, but it’s still $249 — so you should only choose this console if you really like the layout (and don’t mind skipping a screen).

Which Loupedeck is the Best Choice for You?

If you do any streaming or want the deck for general productivity work, the Loupedeck Live and CT are for you. The Loupedeck+ is mainly for professional photo and video editors who will use its dials for the purposes they have printed on them.

The Loupedeck Live and the Loupedeck CT integrate with with Twitch, Adobe Suite, OBS, Spotify, and about 70 other programs, while the Loupedeck+ supports Lightroom Classic, Premiere Pro, Final Cut Pro, Photoshop with Camera Raw, After Effects, Audition, and Aurora HDR. They’re all highly customizable and can control volume and launch any shortcuts you want. If you want to create a shortcut that a program doesn't have, you can use the Loupedeck software to create a custom command.

The consoles also have the ability to save various profiles, either for different use cases or different users. The touch screen ability on the Loupedeck Live and CT allow you to swipe between screens or select actions, such as opening OBS. All models have Dynamic Mode, which means they can switch button functions and screens automatically as you move between different programs (e.g. OBS and Photoshop).

Another satisfying feature of the Loupedeck consoles is the haptic feedback you get as you touch the screen, rotate the buttons, or press in the buttons. The dial on the CT rotates smoothly, with no notches, no feedback, and no push in feature. The dial also has a mini LED screen so it’s unexpected to have multiple functions. Even the Control Dial on the Loupedeck+ has notches and pushes in creating another button function. On both the Loupedeck CT and Loupedeck+ the square buttons seem to contain blue switches.

For streamers and general productivity users, the Loupedeck Live is the best choice, while the Loupedeck CT is best for streamers who also do some video editing. The Loupedeck+ is for heavy video and photo editors.

I prefer the Loupedeck CT, as a content creator who streams and also dabbles in video editing. I’m mainly a writer, so I can setup profiles to streamline my writing process — the same goes for video editing. I’m an intermediate-level video editor, so the programmable Loupedeck shortcuts help a lot as I don’t yet have my video editing app’s menu and keyboard shortcuts memorized.

The Loupedeck consoles don’t cut my writing time in half, but they do cut down on how much time I spend digging through software menus, which is very nice. The Loupedeck+ is a different beast that’s somewhat above me — I don’t consistently edit photos or videos at that level, but it’s nice to know the options are there if I ever want to go deeper.

Despite their status as premium and often expensive displays, the 4K category is red hot if you can take the sheer number of product introductions as an indicator. 32-inch monitors are especially popular, with multiple models landing on shelves and desks from every major manufacturer.

Corsair has been around for nearly 30 years, but it's only recent delved into the display market. I’ve already reviewed the Xeneon 32QHD165 and found it fulfilled its promise as a premium gaming monitor, so I went into this review with high expectations.

The Xeneon 32UHD144, as the name indicates shockingly clearly, is a 32-inch 4K monitor that runs at 144 Hz. The Fast IPS Quantum Dot panel sports Adaptive-Sync, HDR600 and a wide color gamut. It also accepts a line of addons like webcams and peripheral controllers that integrate into the stand to create a hardware ecosystem. Let’s take deep dive into its features and tested performance to see if the Xeneon 32UHD144 is a worthy addition to our best 4K gaming monitors list. 

 Corsair Xeneon 32UHD144  Specs

Corsair hasn’t cut any corners with the 32UHD144. It's high-end in every respect. The Fast IPS Quantum Dot panel has a huge color gamut that qualifies as a Rec.2020 display. It covers over 83% of that gamut, which means it exceeds DCI-P3 by a healthy margin. I measured just shy of 117%. This is one colorful monitor.

Accuracy is provided in equal amountse. There are preset modes for Adobe RGB, DCI-P3 and sRGB, all of which aced my color gamut tests. I recorded numbers that rival the best and most expensive professional screens. There’s also plenty of flexibility, as you can calibrate each mode with its own custom color temp settings. That’s something I rarely see in any display.

Picture quality is further enhanced with an edge backlight with 16 dimming zones. You can use the dynamic contrast feature for both SDR and HDR content and see contrast ratios over 10,000:1. The 32UHD144 is also DisplayHDR 600 certified, which I confirmed in my tests.

Of course, gaming performance is important as well. To that end, Corsair provides a dynamic overdrive that varies its operation based on frame rate for more effective blur reduction. You get Adaptive-Sync with both FreeSync and G-Sync operation confirmed in my tests. The 32UHD144 has not been certified by Nvidia. A backlight strobe feature is also included that can be used instead of Adaptive-Sync.

Corsair has kept the same hardware ecosystem I saw in my review of the 32QHD165. The stand has a threaded fitting called Multi Mount that can accept peripherals like webcams or lights using parts from Elgato. Controllers like Stream Deck and iCue Nexus can be used to quickly change settings without visiting the OSD. It also features the latest video interfaces like DisplayPort 1.4, HDMI 2.1 and USB-C. 

Assembly and Accessories

The Xeneon 32UHD144 comes in a generous box with plenty of crumbly foam to protect its contents. You’ll have to reach for a Phillips-head screwdriver to attach the panel to the upright. It then mates to the substantial metal base with a captive fastener. The cable management clips are already in place and slide up and down to keep the wiring tidy. The cable bundle includes a large external power supply plus USB-C-to-C and A-to-C, HDMI and DisplayPort. 

Product 360

You’ll be hard pressed to tell the difference between a 32QHD165 and a 32UHD144 when there is no image on the screen. The two monitors are physically identical, both panel and stand. The bezel is a thin, flush-mounted affair with a narrow 9mm frame around the top and sides and 20mm on the bottom. A small Corsair logo, the familiar sail graphic, and the name Corsair on the base are the only visible branding.

The 32UHD144’s stand is one of the best I’ve seen. The base is solid aluminum with a gray satin finish. The upright is mostly metal and rock solid. The entire chassis is monolithic in both look and feel with firm adjustments. You get 110mm of height along with 5/20 degrees tilt and 30 degrees swivel. There is no portrait mode. The panel has a 100mm VESA mount in back if you prefer to use your own hardware. Corsair will sell you the monitor without a stand for a $100 discount.

Peripheral options are something unique to the Xeneon line of Corsair displays. At the top of the stand is a threaded fitting that accepts arms from Elgato, which are sold on Corsair’s website. You can attach anything with a standard tripod thread like webcams, lights or microphones. You can also hook up a Stream Deck or iCue Nexus and create macro buttons to change monitor settings on the fly.

In the back, you’ll find a set of cable clips on the upright that slide up and down to help manage the wiring. Here too, you can see the power button and joystick, which controls all monitor functions. There is no LED lighting, nor are there integrated speakers.

The input panel has all the latest interfaces. Video can feed through a DisplayPort 1.4, two HDMI 2.1, or a USB-C connection. A second USB-C upstream port is provided for peripherals, for which there are also two downstream Type-A ports. A 3.5mm headphone jack is provided as well.

OSD Features

The 32UHD144’s OSD is all business with a text-based system divided into six sub-menus. The only graphic is Corsair’s sail logo at the top. It’s summoned with a press of the joystick.

The first menu, Picture, has everything needed for image control and calibration. Standard mode is the default and out of the gate, it’s very accurate. It uses the 32UHD144’s full color gamut, which covers over 116% of DCI-P3. There are eight total modes which include gamut-specific settings for DCI-P3, Adobe RGB and sRGB. They are also extremely accurate.

The Color Temp menu has three presets plus a Custom mode specific to each picture mode. The 32UHD144 is a rare monitor with separate memories for each picture mode. That means you can create a custom calibration for each mode. I’ll explain this in more detail later, but it’s something that even some professional screens don’t offer. You also get three gamma options ranging from 2.0 to 2.4.

The overdrive has three speeds plus a dynamic option which is very effective. It varies the amount of overshoot based on frame rate. I found it cleaned up blur nicely without any ghosting. The fixed settings do create ghosting, so I avoided them. You also get MPRT which is a backlight strobe option. It operates instead of Adaptive-Sync and cuts brightness by around 30%.

The 32UHD144’s backlight is an edge array with 16 dimming zones. You can turn on the dimming for both SDR and HDR content and up the static contrast ratio to around 10,000:1.

Gaming enhancements include a frame rate counter and a set of aiming points that come in red or green with cross, dot or angle shapes.

My only nitpick with the OSD is that to turn on MPRT, you have to go into System Setting and turn off Adaptive-Sync first. This seems like a needless extra step when the AS toggle could have been added to the picture menu. The final Info screen shows resolution, refresh, HDR and Adaptive-Sync status along with the active input and the firmware version. 

Corsair Xeneon 32UHD144 Calibration Settings

There are a number of ways to optimize the 32UHD144’s picture for maximum accuracy. Out of the box, color is very close to the Rec.2020 spec, but grayscale runs slightly warm in tone. Gamma is super tight no matter what options you choose. If you switch to one of the specific gamut modes, Adobe RGB, DCI-P3 or sRGB, you get very high accuracy. The best part is the custom color temp sliders can be set independently for each mode. I’ve done this for you, and my settings are in the table below.

HDR signals lock out all picture controls, but that mode is super precise with excellent grayscale and EOTF tracking. You can see those test results on page five.

Gaming and Hands-on

The Xeneon 32UHD144 is a supremely useful display that serves equally well for work or play. As a productivity tool, it has all the attributes to help power through documents or do research, or do both simultaneously. Or you can keep a news crawl going in a corner while you watch a video. Big monitors like this can be used in a multitude of ways, and the flat panel means no image distortion. 32-inches in the 16:9 aspect is large but not overwhelming. Anything larger like a 43-inch, or wider like a 38-inch 21:9 screen and you’ll have to make some decisions about desktop space.

The picture is razor sharp like any premium 4K screen, but the color stands out. There are a few other 32-inch 4K monitors that have similarly large color gamuts, which I’ll cover in the test comparisons but suffice to say that the 32UHD144 is very colorful. You won’t find a larger gamut in this price range or in the very expensive area occupied by FALD and Mini LED panels.

I saw no benefit to running HDR for Windows productivity apps. Things like Word or Excel don’t look any different. If you want to edit HDR graphics though, it’s an easy switch in Control Panel. Though this is an HDR600 panel and thus very bright, it isn’t too harsh to show pictures in Photoshop or a graphics viewer. HDR streams from Netflix and Paramount+ shine with bright highlights, deep blacks and vibrant color. If you want to watch TV on the 32UHD144, it is more than up to the task.

There is no weakness in gaming performance here. Adaptive-Sync worked perfectly on my FreeSync and G-Sync testbeds with HDR and dynamic overdrive. This feature is one of the things worth buying the 32UHD144 for, and it is not available on many other screens. However, I didn’t see any benefit to the backlight strobe. I had to turn off Adaptive Sync, which at 100-120fps isn’t ideal, and it caused some ringing around moving objects. There were no artifacts when playing with G-Sync and dynamic overdrive, with speeds in Doom Eternal hovering around 120fps.

The 32UHD144’s low input lag definitely made a difference in my personal play experience. I play the same set of games for each review to make fair comparisons between displays and in this case, I had greater success in Doom Eternal’s Horde mode. I also had an easier time dispatching enemies in Call of Duty WWI and Tomb Raider. This is a very addicting display. I expect console gamers will be equally enamored since it’s ready for the 120 Hz VRR with HDR coming from the latest Xbox and PlayStation systems.

The rendered HDR image is stunning. I tweaked Doom’s internal picture controls and was rewarded with a foreboding and hellish landscape that burst with texture. The difference between artificial and natural surfaces was clear. This is an area where 4K is distinctly better than QHD, as long as you don’t move the mouse too quickly. Once the action speeds up, QHD at 165 Hz will look as good or better. However, knowing this did not reduce my enjoyment of the 32UHD144. It is a very capable and amazing-looking gaming screen. 

The comparison group for the Xeneon 32UHD144 is all 32-inch 4K IPS screens running at 144 Hz. We have Philips’ 329M1RV, Dell’s G3223Q, MSI’s MPG321UR-QDX, BenQ’s EX3210U and the Mini LED ViewSonic XG321UG. That last one costs around twice as much as the others. It’s here to provide a frame of reference for what you get when you buy a top-of-the-line 4K monitor. 

Pixel Response and Input Lag

Click here to read up on our pixel response and input lag testing procedures.

144 Hz means a 7ms screen draw time in all cases but the MSI. That one has a tiny advantage with slightly less motion blur at full speed. The 32UHD144 is in line with the others for sheer speed. It is the only one with variable overdrive, which helps up motion resolution a bit over a monitor with fixed overdrive.

It is also one of the quickest overall 4K screens I’ve measured. 30ms isn’t too far behind the capabilities of a QHD/165 Hz monitor. Though the 165 Hz screen will have smoother motion, control response and feel will be similar. The 32UHD144 is a very worthy gaming monitor that will satisfy a player’s wish for high pixel density and low input lag.

Viewing Angles

The 32UHD144 has an off-axis image typical of IPS monitors. At 45 degrees to the side, you’ll see a 30% light reduction and a shift to reddish green. The detail in shadow areas will be a little harder to make out. You can see that the 0 and 10% steps look the same in the photos. The top view is washed out with a blue tint and a 40% reduction in brightness. 

Screen Uniformity

To learn how we measure screen uniformity, click here.

The 32UHD144 does well in the uniformity test with no visible glow or bleed. The light meter detected a little extra light in the lower left corner, but this wasn’t enough to be seen with the naked eye. This is clearly a quality panel built to a high standard.

To read about our monitor tests in-depth, please check out Display Testing Explained: How We Test PC Monitors. We cover brightness and contrast testing on page two.

Uncalibrated – Maximum Backlight Level

The 32UHD144 has no trouble exceeding its brightness spec of 400 nits in SDR mode. There is plenty of light here for any environment, including outdoors. The downside of this is a relatively narrow output range. I couldn’t get the backlight down to 50 nits for dark room gameplay. The minimum is 81 nits which is a little fatiguing if you plan to play for hours in the dark. Native contrast is an IPS average 1,050.3:1. This increases to around 10,000:1 when using the local dimming option in SDR mode.

After Calibration to 200 nits

After calibrating the Standard picture mode, contrast drops slightly to 963:1. There is no visible reduction in picture quality. There is so much color that you almost don’t notice the black levels. The image is richly saturated and detailed.

ANSI contrast is around the same level at 935.5:1. The 32UHD144 is a well-engineered and carefully built monitor with premium components and quality control. 

The 32UHD144 can be optimized in many ways using the preset color gamut modes. Standard uses the full native color space, which is close to Rec.2020. Alternately, you can select Adobe RGB, DCI-P3 or sRGB and calibrate them individually. 

Grayscale and Gamma Tracking

Our grayscale and gamma tests use Calman calibration software from Portrait Displays. We describe our grayscale and gamma tests in detail here.

Grayscale tracking is very consistent across all picture modes. Out of the box, the 32UHD144 runs a little warm in tone, but the errors are slight and only visible above 60% brightness. Gamma tracking is about as close to perfect as a monitor can get.

Calibrating the custom color temp removes all visible errors and takes the 32UHD144 to a reference level of accuracy. Gamma gets slightly lighter on the chart, but this change cannot be seen with the naked eye. I was able to achieve the same results in every gamut mode, so I’ve only included a single before and after chart here.

Comparisons

With a default grayscale error of 3.23dE, the 32UHD144 can be enjoyed without calibration. But since there is such upward potential, it makes sense to tweak a bit. I got the average error below 1dE in all gamut modes using independent RGB settings. You’ll find them in the table back on page one. The post-calibration errors for the other modes are DCI-P3 – 0.63dE, Adobe RGB – 0.58dE and sRGB – 1.06dE.

Gamma tracking is among the best in class. The range of values is super tight at only 0.06 from lowest to highest. The average value is 2.16 for a 1.82% deviation, which is also excellent. This holds true in all gamut modes.

Color Gamut Accuracy

Our color gamut and volume testing use Portrait Displays’ Calman software. For details on our color gamut testing and volume calculations, click here.

I’ve included charts showing all the 32UHD144’s gamut modes. The default measurements for Standard mode, which are set against the Rec.2020 standard, show some hue errors in red and magenta and slight under-saturation overall. Note that this is under-saturation in the Rec.2020 realm, DCI-P3 would show a general over-saturation.

If you pick a specific gamut mode, the 32UHD144 hits the marks right on the nose. You can see that every point is on target for DCI-P3, Adobe RGB and sRGB. It doesn’t get better than this.

Comparisons

I’ve charted the error for Standard mode, which uses the 32UHD144’s full native gamut. The other color modes have even lower values. Adobe RGB is 1.70dE, DCI-P3 is 1.32dE and sRGB is 1.33dE. That result exceeds many professional monitors that cost substantially more.

The 32UHD144’s near-117% coverage of DCI-P3 should be no surprise, given that it covers 83.36% of Rec.2020. That makes it one of the most colorful monitors I’ve tested. You can get similar coverage from the BenQ, MSI and ViewSonic screens with comparable levels of accuracy.

Our HDR benchmarking uses Portrait Displays’ Calman software. To learn about our HDR testing, see our breakdown of how we test PC monitors.

The 32UHD144 is a very capable HDR monitor with a zone dimming edge backlight. It has 16 dimming zones, which enhance contrast significantly. An HDR10 signal switches modes automatically. 

HDR Brightness and Contrast

The 32UHD144 is rated for 600 nits in HDR mode, and my sample beat that handily with nearly 700 nits of total output measured using a full-field white pattern. That’s seriously bright. In practice, you’ll only see this level in small highlight areas. Thanks to the effective zone dimming feature, black levels are relatively low, and contrast is above average at 10,875.2:1. The Mini LED ViewSonic achieves high values thanks to its incredibly bright backlight and super low black levels. But that level of performance will cost you about double the price of the Corsair and the other screens.

Grayscale, EOTF and Color

I was surprised to see a lack of image control in HDR mode. There is only a single picture mode and no means of changing color. But obviously, that isn’t a problem given the results shown above. The 32UHD144 is one of the most color-accurate HDR monitors I’ve tested. Grayscale tracking is visually perfect. The only flaw is some lightness at the lowest level. This is likely due to the speed in which the zone dimming backlight reacts to changes in picture levels. In practice, the HDR image is very deep with true blacks and very bright highlights.

In the color test, the 32UHD144 is clearly using the Rec.2020 reference for all content, even when mastered to DCI-P3. You can see in the DCI chart how green tracks toward the Rec.2020 points, which means it’s a tad bluer. This will be hard to spot in actual content, but a firmware update might be in order to ensure that color tracks correctly for the content’s reference gamut. Saturation points are linear in progression with slight over-saturation overall, not a bad thing.

In the HDR monitor category, there are three distinct tiers of image quality. Budget screens make do with no dynamic contrast and therefore have images that look much the same whether they’re SDR or HDR. At the high end where prices can top $2,000, we have full-array local dimming backlights, some with Mini LEDs, and potentially over 1,000 dimming zones. They look stunning but their prices are high.

In the middle, for $800-1,000, are monitors like the Corsair Xeneon 32UHD144. They deliver plenty of brightness and zone dimming edge backlights to get HDR contrast over 10,000:1. That’s a significant difference from the IPS typical 1,000:1. And those monitors have color gamuts just as large, if not larger, than their more expensive cousins. They are in the sweet spot of the price/performance ratio.

The 32UHD144 checks all the boxes for a premium 4K screen. Its video processing is excellent thanks to dynamic overdrive, something very few gaming monitors offer. The image is stunning thanks to that Quantum Dot film and the resulting huge color gamut. With nearly 117% coverage of DCI-P3, it qualifies as a Rec.2020 panel. Again, only a few other screens have this much color.

Corsair also heads the class in color accuracy. With preset picture modes for Adobe RGB, DCI-P3 and sRGB that are all spot-on, you can use it for the same color-critical work normally taken on by a much more expensive professional display. The 32UHD144 is a reference-level monitor.

While the price at this writing is $999 ($899 without the stand), it doesn’t cost much more than other comparable products. And its color accuracy, build quality and ability to anchor a hardware ecosystem make it stand out. The Corsair Xeneon 32UHD144 is a monitor anyone looking to go 4K should strongly consider.

Today, in an Xbox news post, Microsoft announced a new image-enhancing feature coming to the Edge browser called Clarity Boost. The feature adds clarity to games streamed from Microsoft's own Xbox Cloud Gaming Servers. The service will be exclusive to Xbox's streaming service alone, with no word on Microsoft if the feature will extend to other platforms as well.

The feature is available to try now through the Microsoft Edge Canary Channel, which you can check out here. Clarity Boost will arrive officially to Edge sometime in 2022.

To demonstrate the feature, Microsoft has provided an image with both Clarity Boost on and off. The feature works very similarly to the clarity sliders you might find in Adobe Lightroom or Photoshop, in that it adds a bit of contrast to areas of the image with sharp edges. This makes the image sharper and reduces blurring for an overall cleaner image.

Microsoft seems to have tuned the Clarity Boost option very well, it's neither too low nor too high, and it really does make the image look a bit better. In addition, the adjustment is available as a toggle when you start streaming Xbox games from the Edge canary browser.

But strangely, the feature will be exclusive to the Edge browser and it doesn't appear to be making its way into the Xbox app for PC, which is another way you can stream Xbox games from PCs. We're not sure why this is the case, but possibly Microsoft is testing this feature out before it makes its way into the Xbox app or even Xbox consoles.

For full details on Xbox's Cloud Gaming streaming service, check our coverage here, and if you want to try streaming Xbox games from a Rasberry Pi, we have a tutorial you can check out here as well.

Hardware detective Tum_Apisak has dug up new benchmarks for Intel's forthcoming Core i5-12400.  It's an unreleased chip, but a eBay seller recently put the hexa-core Alder Lake processor up for purchase for $364.77.

It's not the first time that the Core i5-12400 is making rounds in the hardware circles. Last month, a Bilibili user shared some Core i5-12400 benchmarks where the chip was putting up some very strong performance. On this occasion, we get to see the Core i5-12400 go up against the Ryzen 5 5600X in Adobe Premiere Pro from the PugetBench benchmark suite.

If we put all the small leaks together, the Core i5-12400 reportedly comes equipped with six Golden Cove cores at 2.5 GHz that can boost up to 4.4 GHz. The Alder Lake processor also sports 18MB of L3 cache and a 65W PBP (Processor Base Power). The Ryzen 5 5600X, on the other hand, wields six of AMD's Zen 3 cores that operate with a 3.7 GHz base clock and 4.6 GHz boost clock. The chip has 32MB of L3 cache and a 65W TDP. The Ryzen 5 5600X lacks integrated graphics so the Core i5-12400F is a more fair comparison. However, it doesn't matter since Intel's F-series SKUs have the same level of performance as their non-F counterparts.

Intel Core i5-12400 Benchmarks

As with all benchmark leaks, it's useful to sprinkle some salt over them. Although Intel has likely already shipped the Core i5-12400 to OEMs, there's no way to know whether the Core i5-12400 from the benchmark is a retail sample or not. There are some details that may affect the Core i5-12400's numbers, though. For starters the Core i5-12400 was paired with DDR4-2400 memory whereas the Ryzen 5 5600X had access to faster DDR4-4000 memory. Both systems use the GeForce RTX 3060 Ti and Windows 10 21H1 (19043).

Windows 11 does give Alder Lake a nice performance boost in some workloads. Our Alder Lake review shows that Alder Lake and Ryzen 5000 perform similarly in Premiere Pro regardless if it's Windows 10 or Windows 11. Therefore, we don't expect the operating system to have hindered the Core i5-12400's performance.

Overall, the Ryzen 5 5600X delivered just 1.4% more performance than the Core i5-12400. The delta is so small that it's within the margin of error. Heck, you could even argue that the Core i5-12400 might actually be faster if was running with the same DDR4-4000 memory. Unfortunately, that's something we won't know until Intel officially launches its Alder Lake non-K series processors.

The very first benchmark results of Apple's M1 Max system-on-chip in a professional application have arrived. While performance of the new M1 Max-based MacBook Pro in Adobe Premiere Pro looks very good compared to the previous-generation MBP with discrete graphics, it doesn't look that good compared to x86 workstation platforms with standalone graphics processors.

On Thursday an M1 Max-based MacBook Pro user posted performance numbers obtained in Adobe Premiere Pro to the PugetBench database. Benchmark results were obtained using PugetBench for Premiere Pro 0.95.1 and Premiere Pro 15.4.1, so it is possible to compare them to results obtained on other high-performance systems and find out how the new MBP stacks up against machines based on x86 CPUs and discrete GPUs.

Adobe's Premiere Pro 15.4.1 video editor is one of the programs that can take advantage of virtually all kinds of processing units, including general-purpose CPU cores, GPU cores, and media playback engines. Therefore, the application can demonstrate capabilities of a platform from different aspects. PugetBench for Premiere Pro measures both live playback and export performance with a wide range of codecs at 4K and 8K resolutions. Furthermore, there are dedicated "Heavy GPU Effects" and "Heavy CPU Effects" sequences in Export tests designed to stress appropriate hardware. 

Overall

Apple's MacBook Pro 16 based on the M1 Max SoC achieved a 1168 standard overall score and 1000 extended overall score in PugetBench for Premiere Pro 0.95.1. Significantly higher when compared to high-end laptops. Furthermore, Apple's scores are close to those of advanced desktops. 

For obvious reasons, PugetBench for Premiere Pro 0.95.1 (which uses Premiere Pro 15.4.1) calculates its overall score based on various kinds of workloads, therefore it makes sense to take a look at detailed results and see where exactly Apple's new SoC excels. 

Live Playback

Apple's M1 Max clearly has an excellent media playback engine that outperforms not only standalone mobile GPUs, but even Nvidia's top-of-the-range GeForce RTX 3090. Those who edit video will certainly appreciate Apple's M1 Max since based on PugetBench for Premiere Pro 0.95.1, the new MacBook Pro systems promise a very smooth experience that will be even better than that on beefy desktops.

Export

The new MacBook Pro 16 clearly outperforms its predecessor in Export workloads, but is somewhat behind other high-end laptops and is drastically behind advanced desktops.  

It is necessary to point out that PugetBench for Premiere Pro's Export workloads include heavy CPU effects and heavy GPU effects sequences, so overall score is somewhat mixed. It still makes a lot of sense because real-world workflows may require different types of effects.  

We suspect that Apple's M1 Max outshines the competition in heavy CPU effects sequences, but falls behind in heavy GPU effects sequences, so it really depends on the exact workload whether the new MBPs provide a better experience than other platforms or cannot keep up with the rivals.

GPU Score

Adobe's Premiere Pro is one of those professional applications that can take advantage of GPU compute capabilities, so it is not surprising that the new MacBook Pro with its 32-cluster custom GPU with 4096 ALUs annihilates AMD's Radeon Pro 5500M used in previous-generation MacBook Pro workstations. 

The new M1 Max SoC can also compete very well against standalone mobile GPUs, namely the GeForce RTX 3060 and RTX 3080 (which seems strangely slow in this benchmark), in Premier Pro while consuming much less power. But Apple's new integrated GPU cannot get close to performance levels offered by desktop discrete graphics cards, something that Apple needs for its Mac Pro workstations.

Some Thoughts

When Apple introduced its M1 Pro and M1 Max earlier this week, it did not demonstrate their performance in any specific workstation application, but focused on performance and power consumption in a host of programs and compared its CPU and GPU performance to that of other notebooks. While we can expect the new SoCs to offer a good balance between performance and power consumption, just like M1, specifics are important. Since Apple has not provided them, we have to fetch such information from other sources, which is not 100% accurate, but this is all we have at this point.

Apple clearly positions its M1 Pro and M1 Max system-on-chips for workstation applications, so while it is fun to see how the M1 Max SoC compares to CPUs and discrete GPUs in things like GFXBench 5 or Geekbench 5, these are certainly not workloads that this chip was designed for. 

Adobe's Premiere Pro 15.4.1 is one of the programs that will be used on the new MacBook Pro laptops and performance in this program shows that Apple's latest M1 Max SoC has an industry-leading media playback engine, state-of-the-art general-purpose CPU cores, and a GPU that can outrival standalone mobile GPUs from Nvidia in Premiere Pro. Nonetheless, Apple yet has to develop a GPU that will offer performance on par with discrete desktop GPUs from Nvidia.

The end has finally come for Adobe Flash on the Raspberry Pi. What once powered many memes and early Internet games is no more. With the latest update to Raspberry Pi OS, Adobe Flash has been removed from the official Raspberry Pi operating system but this brings about an issue of its own.

Adobe Flash was once the darling of the late 1990s and early 2000s web. Embedded games, interactive interfaces and full blown applications were built with Flash, but over time it became bloated, buggy and a security risk. The removal of Flash from Raspberry Pi OS has been on the cards for quite some time and with the January 11 update they have finally removed it. The removal of Flash does have one impact on Raspberry Pi OS. The popular Scratch 2 programming application will now no longer work as it relies upon Flash. So before you update your Raspberry Pi to the latest OS, ensure that your projects work with the Scratch online editor.

Looking through the release notes, the other changes to Raspberry Pi OS in this release are largely bug fixes for the new PulseAudio configurations introduced in the December 2020 update. There are a few fixes for the Chromium web browser, notably removal of artifacts in Google Maps 3D view.  An updated Linux Kernel, now 5.4.83 and updated Raspberry Pi firmware keeps your Pi securely up to date. 

Raspberry Pi OS is available from the Raspberry Pi website, where it can be downloaded as a disk image, or via the Raspberry Pi Imager tool.

Microsoft this week finally released a version of Windows 10 on Arm that supports x86-64 (x64) emulation to members of the Windows Insider program. The software giant says that the preview version of Windows 10 on Arm supports any x64 app whether they come from the Microsoft Store or any other source. 

Members of Microsoft's Windows Insider program with PCs powered by Qualcomm's Snapdragon (or any other) ARM64/Armv8 system-on-chips (SoCs) who want to try the new version of Windows on Arm with x64 emulation should download the Windows 10 Insider Preview Build 21277 (RS_PRERELEASE) from the Dev Channel.  

To get decent app performance, users will also need to install a proper x64 driver for their Qualcomm Adreno GPUs, which is currently available only for Samsung's Galaxy Book S, Lenovo's Flex 5G, and Microsoft's Surface Pro X (more information and links here). Also, to support running both ARM64 and x64 C++ applications simultaneously, Microsoft advises installing a preview version of the ARM64 C++ redistributable. 

Microsoft does not reveal any information about its x64 emulator's expected performance for Armv8 processors, so it is unclear whether the software is good enough for demanding applications like Adobe's Photoshop, or is mostly aimed at things like browsers. It is also unclear how stable the emulator is. Meanwhile, the software giant recommends developers support their apps natively on ARM64 and has been providing appropriate tools — Windows SDK and Visual Studio — since 2017.  

Microsoft has been actively trying to bring Arm SoCs to the Windows PC space for almost a decade. Initially, the company wanted to offer an alternative to Apple's iPad. It released its Windows RT operating system in 2012, but the platform has never been popular, and Microsoft essentially canned the project in 2015 by axing its own Windows RT devices. The second attempt — sometimes called always-connected personal computers (ACPS) — was unveiled by Microsoft and Qualcomm in 2017. Instead of bringing an alternative to something on the market already, the two companies focused on a unique value proposition that Snapdragon SoCs can provide: long battery life and always-on connectivity. 

In addition to launching an x64 emulator for Armv8 SoCs, Microsoft also plans to release a number of its programs, including Microsoft Teams, compiled and optimized specifically for Arm SoCs. Furthermore, Microsoft is also expanding its App Assure program to include Windows 10 on Arm 64-bit devices to guarantee that custom software designed for particular customers would work on all Windows 10 platforms no matter which processor or SoC they use. 

Today, the Raspberry Pi Foundation announced an end of year update to its 32-bit Raspberry Pi OS, formerly Raspbian. This update sees improvements to Chromium, PulseAudio becoming the default audio server, and printing is made much easier. The update also provides configuration options for the Raspberry Pi 400 and the new Raspberry Pi 4 Case Fan.

Starting with improvements to the Chromium web browser (now version 84), thanks to the Raspberry Pi team's work, we finally see hardware-accelerated video playback for YouTube videos and streaming media services. The team also improved video conferencing clients such as Google Meet, Zoom, and Microsoft Teams, a welcome addition given the world that we now live in. The final note for Chromium is that version 84 of the browser will be the last to come with Adobe Flash Player. Adobe will retire the player at the end of 2020.  

The PulseAudio audio server replaces the long-standing ALSA server. With PulseAudio, we see the ability to play audio from multiple sources at once and native support for Bluetooth audio (unlike ALSA, which uses a third party bluez-alsa package to achieve this). 

Simon Long, Senior Principal Software Engineer, explains that the transition to PulseAudio should be seamless for the end user, "The good news for Raspberry Pi users is that, if we’ve got it right, you shouldn’t even notice the change. PulseAudio now runs by default, and while the volume control and audio input/output selector on the taskbar looks almost identical to the one in previous releases of the OS, it is now controlling PulseAudio rather than ALSA. You can use it just as before: select your output and input devices, adjust the volume, and you’re good to go."

Printing in Linux is handled via CUPS, the Common Unix Printing System. With this latest update, we see CUPS pre-installed along with a printer config GUI tool to configure your printer over USB or a network connection easily.

Raspberry Pi 400 and Raspberry Pi 4 Case Fan Updates

The Raspberry Pi 400 was quite a surprise, and we loved the new form factor, but one issue that it shares with the Raspberry Pi Zero range of boards is a single status LED to show power and "drive" activity. With the latest Raspberry Pi OS update, we see an option in the Raspberry Pi Configuration tool that allows users to choose between a simple power LED or to have the LED show drive activity.

A further update to the Raspberry Pi Configuration tool also sees an extra option in Performance for the new Raspberry Pi 4 Case Fan. With this option, we can set the temperature at which the fan is triggered and set the GPIO pin used to control the fan.

The latest version of Raspberry Pi OS can be downloaded directly from the Raspberry Pi website. If you have the Raspberry Pi Imager tool, you already have the update ready to download and flash to a micro SD card. If you would like to update an existing install, there are full instructions on the Raspberry Pi website.

Logitech is showing Mac workhorses some love. Today, it released a version of the Logitech MX Master 3, our favorite productivity mouse ever, that works with Apple computers and tablets. That means the mouse can control and share files among three Mac desktops, laptops or an iPad and launch profiles catered to software made for Apple products. 

We asked Logitech if the MX Master 3 for Mac, as well as the MX Master Keys for Mac keyboard would still work with full functionality once software comes out that is redesigned for Apple Silicon, Apple’s homegrown CPUs that will start taking over for Intel chips starting at the end of this year. However, the vendor told Tom’s Hardware that it didn’t have any information on that yet, since the announcement is very new. 

Logitech MX Master 3 for Mac

The Logitech MX Master 3 for Mac is available today for $99.99, the same price as the Windows version, but optimized for MacOS and iPadOS. The wireless mouse comes in a “Space Gray” meant to mimic the hues of Apple devices, although it seems a little dark from our perspective.  

This is a Bluetooth wireless mouse that’s supposed to last for up to 70 days without a charge. While the PC version of the MX Master 3 includes a dongle, Logitech ditched the dongle connection with the Mac version because it believes Mac users are more likely to use a Bluetooth connection than a dongle.  

The Logitech MX Master 3 for Mac works with free Logitech Options software, which comes with preset profiles that have the 6 programmable buttons set with functions for popular apps, like Adobe Photoshop, Final Cut Pro and Google Chrome. But for the Mac version, Logitech fine-tuned for the Mac versions of these apps and included programs like Safari. The mouse is programmed to appeal to advanced content creators, but you can also edit and create your own profiles in Logitech Options. 

As we detailed in our Logitech MX Master 3 review, the mouse is a multi-PC productivity machine, allowing you to control and share files, text and photos across up to three PCs. That’s thanks to Logitech’s Flow feature that works similarly to Windows 10’s Nearby Sharing.

What really stands out with this mouse is its scroll wheel, dubbed MagSpeed. It’s a different wheel than what’s in the MX Master 2S and MX Master. The reinvented wheel uses two electromagnets that allow you to switch between smooth, super-fast scrolling or more precise ratchet movements. Unlike its predecessors, the MX Master 3 does each type of scroll silently, and its free spin is as good as it gets when it comes to speed. 

Logitech designed the mouse for palm grippers, and you get a side thumb wheel that’s actually accessible and comfortable. 

With this mouse still being my go-to productivity mouse after reviewing it in November, it’s only fair that Mac users can now consider it too. With a Logitech Darkfield laser sensor sporting 4,000 CPI (counts per inch), it’s not meant to be the best gaming mouse, but instead a reliable and sturdy mouse for your toughest workloads. 

Logitech MX Keys for Mac 

Today also brought the arrival of the Logitech MX Keys for Mac. This is an Apple-friendly version of the Logitech MX Master Keys that also offers wireless connection (Bluetooth or dongle) and control and file sharing among three PCs. You can also program the function row to perform different actions with different Apple software.

But since it targets Mac users, Logitech updated the keyboard with a Mac layout. And while the eject button is only functionable on its own with an iPad (it shows or hides the keyboard), Logitech added hot keys to make the button work with others to make your system go to sleep, restart, turn off or put a display to sleep while the Mac is awake. 

The Mac version of this keyboard also comes with a USB-C to USB-C charging cable, instead of the USB-C to USB-A one that comes with the Windows version. Logitech said that this is because the latest version of Macs boast as many as four USB-C ports.

With its low-profile membrane keys and build, this won’t make the best gaming keyboard. When we went hands-on with the Logitech MX Master Keys, it was hard to get used to its shallow experience compared to a mechanical keyboard. But in a press briefing, Logitech said that the MX Keys for Mac is meant to remind users of their experience using a Mac keyboard, with a slightly more premium experience than the Apple Magic Keyboard. 

Its backlighting also uses a proximity sensor, so it’ll only light up when you’re near it. And there’s an ambient light sensor, so the MX Keys for Mac can adjust brightness based on the lighting in your room’s environment. 

With Logitech’s MX line of productivity gear finally being fine-tuned for Macs and iPads, Apple fans will have no excuse not to get their work done.  

Apple is transitioning its Mac computers to its own custom chip designs based on Arm. The company is detailing its transition away from Intel's processors at its annual WWDC  (Worldwide Developers Conference). Apple calls the new chip 'Apple Silicon.'

Apple's Johny Srouji, head of the company's chip design, said Apple's CPUs are often more powerful than PC laptops. He said he wants Apple Silicon to be more powerful while consuming less power.

Srouji said the GPU and battery life would be better, and privacy would be paramount. CEO Tim Cook said the company will launch the first Mac with an Apple CPU by the end of the year, while there are still more Intel Macs in the pipeline. He suggested the entire transition will take roughly two years.

The first silicon will be on Macs with macOS 10.16, Big Sur. Apple made every app it has produced to operate natively with the new chips, and developers will merely have to recompile in xCode. Universal 2, a new binary that works on Intel Macs and Apple Silicon Macs, can use the binary for all of their users. Apple's Craig Federighi said Microsoft and Adobe are already working on apps. Those include Word, Excel, PowerPoint, Lightroom and Photoshop.

Federighi said all demos during WWDC ran on a development system using Apple's 12Z Bionic chip.

A new version of Rosetta, Rosetta 2, will let you run apps made for Intel Macs on Apple's own systems. It translates code when apps are installed, but can also do it while you use apps. Apple also showed gaming, including Shadow of the Tomb Raider, working through Rosetta 2 at 1080p with emulation.

Big Sur will have virtualization for Linux and Docker. You'll also be able to run iPhone and iPad apps natively on new Macs, all without developers changing anything.

There will be a Quick Start program for developers, including forums, sample code and access to Apple support. There will also be a hardware developer kit: a Mac Mini with A12Z, 16GB of RAM and a beta of macOS Big Sur.

Analysts have often estimated that Apple's computers make up between 5 to 7% of Intel's client computing business.

"Apple is a customer across several areas of business, and we will continue to support them," an Intel spokesperson told Tom's Hardware. "Intel remains focused on delivering the most advanced PC experiences and a wide range of technology choices that redefine computing. We believe Intel-powered PCs—like those based on our forthcoming Tiger Lake mobile platform—provide global customers the best experience in the areas they value most, as well as the most open platform for developers, both today and into the future."

Apple developing its own chips in-house will allow it further control of its ecosystem, including not tying new computer releases to Intel's schedule. Apple already uses its own processors in the iPad, iPhone and Apple Watch. Its own T2 chip has already taken over some security and SSD controlling processes in its Macs alongside Intel processors, and the company has also made its own GPUs for the iPhone and iPad. Srouji said Apple had shipped over 2 billion SoCs on those devices.

This isn't the first time that Apple has transitioned between processor architectures on the Mac. It moved from the Motorola 68000 series to PowerPC, and then announced a transition from PowerPC to Intel at WWDC in 2005.

Updated June 22 at 3:15 p.m. ET with Intel's response.

Adobe announced three years ago that the company planned to retire Adobe Flash on December 21, 2020. The plan still stands with Adobe making all the necessary preparations for the multimedia software platform's retirement.

During its lifetime, Adobe Flash had been an important part of website, application and game designs. However, it's way past its prime, and Adobe also acknowledges that there are now better alternatives out there, such as HTML5, WebGL or WebAssembly.

Adobe has committed to supporting Adobe Flash with bug fixes, new features and security patches until the end of the year. After that, the company will wheel it into the retirement home and pull all Flash Player download pages from its website. Artists, developers and businesses that didn't plan ahead and replace Flash content will ultimately be left with unusable websites.

You might be tempted to acquire Flash Player from a third-party provider, but Adobe advises against doing so since the company cannot guarantee the authenticity of unauthorized downloads of Flash Player. The best course of action is to just uninstall Flash Player before the EOL (end-of-life) date and move on with life. Adobe will also prompt its users to remove Flash Player from their systems later this year although the company didn't delve into specifics.

We're a little over six months from the end of an era. Let's take this time to appreciate and enjoy what is left of Adobe Flash.

Because there are so many RGB components and accessories, there are just a ton of RGB control panel applications that let you customize your light patterns. However, Thermaltake's new NeonMaker software is something different, because it allows you to create and edit lighting animations in much the same way you'd piece together a video clip.

Available for download right now on Thermaltake's website. NeonMaker allows you to build and save animations that are up to 45 seconds long, using any components that are part of the company's RGB Plus ecosystem. We had a chance to use NeonMaker at Thermaltake's CES 2020 suite and were impressed with its capabilities. 

In the application, all the eligible components appear as icons at the top of the screen and you can drag them around to represent their position in the case.  The setup in our demo was simply a series of Riing Quad case fans so all the icons were the same, but a Thermaltake rep said that other types of coolers such as AIOs would have icons that looked like the products do in real life. 

You can then click on each icon and set its colors. Each RGB device the has its own row in the overall timeline just like each video or audio clip has its own timeline when  you're crafting videos in editors like Adobe Premiere or Final Cut Pro. 

Once you're done with your animation, you can save it to a file, which you can share with friends or reload the next time you want to use it. NeonMaker does not replace Thermaltake's regular RGB Plus lighting control software; it just adds another way to express yourself through your PC's lighting. 

In an incredible alignment of the silicon stars, AMD's third-gen Threadripper 3970X and 3960X launch today alongside Intel's new Cascade Lake-X Core i9-10980XE, giving us a look at the future of the high end desktop all in one go. 

Intel surely has its work cut out for it. Earlier this month, AMD showcased the advantages of its Zen 2 microarchitecture and 7nm process by cramming the 16-core/32-thread Ryzen 9 3950X into its mainstream platforms, doubling the core count of Intel's competing chips. That created a new tier of threaded performance for mainstream motherboards, one Intel simply can't match, and upset Intel's pricing structure for its high-end desktop lineup in the process. 

But that wasn't the full extent of the damage: Intel preemptively slashed the gen-on-gen pricing of its forthcoming Cascade Lake-X chips in half as it braced for the impact of AMD's Threadripper 3000. 

Now they're here: AMD's 32-core 64-thread Threadripper 3970X and 24-core 48-thread 3960X have landed in our labs with the same core counts as their predecessors, but instead of a headline achievement like a doubling of core counts, Threadripper's new architecture serves as the star of the show. The 'Castle Peak' design brings a new design that does away with many of the previous-gen Threadrippers' eccentricities, which equates to massive performance gains across the board. Add in support for PCIe 4.0, and the Threadripper 3000 series looks like a winner.

Of course, with much more real estate available in the massive sTRX4 socket on TRX40 motherboards (), we also fully expect AMD to bring the density advantages of its 7nm process to bear, and soon it will: The company confirmed it would bring a beastly 64-core 128-thread Threadripper 3990X with an amazing 288MB of total cache and 280W TDP to market in 2020. 

But while we await that ludicrously-appointed chip, we have plenty of excitement with the Threadripper 3970X and 3960X. These chips promise a new level of performance for prosumers and creators, and they deliver. 

We've been busy in our labs testing the processors through our standard test suite of games and applications, but also added an expansive set of workstation tests that expose the benefits of the TRX40 platform, like AMD's leading support of PCIe 4.0. We also have Intel's new Core i9-10980XE on the roster, along with the Ryzen 9 3950X, to give you a view of the entire high-end battlefield. 

Threadripper 3970X and 3960X Specifications and Pricing

AMD's Threadripper 3000 processors debut with much higher pricing than the previous-gen models: At $1,999, the 32-core 3970X debuts for $200 more than its predecessor, and AMD tacked on an extra $100 for the $1,399 24-core 3960X. AMD attributes the increased pricing to the new level of performance and access to Threadripper's leading-edge support for the PCIe 4.0 interface. 

AMD gears the Threadripper 3000 processors for content creators and prosumers, like 3D artists, filmmakers, and software developers. For professionals, any extra time spent completing a task equates to lost revenue, making the higher price a justifiable expense if the premium matches the performance.

You'll also need a new TRX40 motherboard to support the chips. We've tested a couple around the launch of these chips, and our current favorite is ASRock's TRX40 Taichi. Due to the higher pricing of Threadripper 3000 models and the requirement for a new motherboard, AMD is leaving the existing second-gen Threadripper processors on the market for the foreseeable future as a 'value HEDT' offering for content creators. You can regularly find the 32-core Threadripper 2990WX for  roughly $1,700 and the 24-core 2970WX for about $925, and we expect those prices will fall soon. 

The Threadripper 3970X weighs in with 32 cores and 64 threads that operate at a 3.7 GHz base and 4.5 GHz boost paired with 144 MB of total cache (128MB L3), while the Threadripper 3960X has 24 cores and 48 threads that run at a 3.8 GHz base and 4.5 GHz boost with 140 MB of total cache. 

Both chips expose 88 PCIe 4.0 lanes, but the TRX40 chipset consumes some of those lanes, leaving 64 exposed to the user. The PCIe 4.0 interface is a notable advantage of AMD's Ryzen desktop lineup, but it is more important in the HEDT space where high-speed storage and networking devices are more likely to find their way into powerful systems. 

Both processors support four channels of DDR4-3200 memory, but data transfer rates differ based on DIMM population. The chips support up to eight DIMMs of 32GB memory, totalling 256GB. The processors also support ECC memory, but implementation varies by motherboard. The Threadripper processors feature the same memory controllers as the Ryzen 3000 chips, so memory overclocking is a rather straightforward affair.

Intel's new Cascade Lake-X represents the company's flagship competitor on the high-end desktop, but it is only $979, which is less than half the price of the previous-gen Core i9-9980XE. That leaves AMD's Threadripper 3000 largely uncontested in the traditional HEDT space, though Intel does offer its Xeon W processors with the server-derived LGA 3647 socket. 

The Xeon W chips slot into a higher workstation tier, top out at 28 cores, and don't feature unlocked multipliers, meaning they aren't overclockable like HEDT processors. These processors require expensive motherboards and coolers, and also feature six-channel memory controllers. Given their workstation-class features, the Xeon W chips that compete directly (based on core count) against Threadripper 3000 retail for $3,349 and $4,499 for a 28-core chip, so they aren't suitable competitors. 

Intel also has its Xeon W-3175X, the lone overclockable processor on the LGA 3647 platform, for $3,000. This 28-core 56-thread processor drops into exotic motherboards that carry heart-stopping price tags, so it really isn't a direct competitor to Threadripper 3000, either.

AMD doesn't specifically market the Threadripper processors for workstation use, and the significantly lower pricing means Intel has largely ceded the top of the HEDT market to AMD, at least for now.

Third-gen Threadripper Architecture

The Threadripper 3000 series chips come with the TSMC 7nm process, which has density advantages that manifest as higher performance, better power efficiency, more cores, and more cache packed into a smaller die area than the first- and second-gen Threadripper models. 

Like the mainstream Ryzen parts, Threadripper 3000 comes packing AMD's Zen 2 microarchitecture that brings a notable IPC improvement, but AMD spreads the design across four eight-core 'core chiplet die' (CCD), as opposed to two with the mainstream chips. AMD ties the compute chiplets together via the Infinity Fabric to a large central 12nm I/O die that houses two 32x PCIe Gen4 controllers and two dual-channel DDR4 memory controllers. 

Each 7nm compute chiplet features ~3.9 billion transistors, while the 12nm I/O die has ~8.34 billion transistors, yielding a total of ~23.94 billion transistors in the Threadripper 3970X and 3960X spread over 712 square millimeters of silicon.

As we can see in the last image in the album below, the new design stands in contrast to the distributed design in the first- and second-gen Threadripper models, which had up to four die connected directly to each other. Unfortunately, AMD fused off the PCIe and memory controllers on two of those die, which led to increased latency when the die needed to access far memory banks and I/O devices. 

Threadripper 3000's I/O die provides the compute chiplets with uniform access to the PCIe and memory controllers, thus reducing a layer of latency and ensuring a smooth quality of service. AMD also reduced the Infinity Fabric On-Package's (IFOP) power consumption by 27%, which the SoC now allocates to provide extra compute horsepower. 

Like all modern processors, AMD's 3000-series processors and dies come with a mix of faster and slower cores. However, as we discovered, AMD programs maximum boost frequencies on a per-core basis. That stands in contrast to the previous industry status-quo of delivering a uniform boost capability across all cores, but the technique extracts the maximum performance and power efficiency out of each die.

As a result, each core and die has different boost capabilities, so AMD chooses a premium die and places it in the CCD4 position inside the SoC. The company then uses an innovative mix of the Windows scheduler, drivers, and motherboard firmware to target the fastest two cores with lightly-threaded workloads, which allows it to execute at the highest possible frequency. 

The Windows scheduler, which interfaces with CPPC2 (aka 'preferred core), rotates the lightly-threaded work between the two fastest cores on CCD4 to reduce current and thermal density, which fosters more frequent and longer-duration boosts. AMD has weathered plenty of criticism for its early missteps when implementing the CPPC2 functionality, but has fixed the lion's share of the issues. As we'll show on the following page, our Threadripper chips had no issues reaching their rated boost speeds, and even exceeded the rating on rare occasions. 

TRX40 Motherboards

AMD's new TRX40 platform features a new sTRX4 socket that isn't backward compatible with existing Threadripper processors, and the 3000 series isn't backward compatible with X399 boards, either. The sockets are mechanically compatible, so you use a cooler designed for previous-gen Threadripper chips with the 3000-series models, and AMD says that any cooling solution capable of handling 1000- and 2000-series processors should suffice for the 3970X and 3960X. 

The sTRX4 socket features the same 4094 pins as the previous-gen sTR4 socket, but AMD changed the electrical layout of the new socket to accommodate the PCIe 4.0 interface. The company hasn't committed to compatibility with future Threadripper processors, but says the socket is designed with scalability in mind. That will obviously come in handy when the 64-core Threadripper 3990X lands next year. 

The TRX40 chipset features 16 GB/s of throughput between the processor and the chipset, which comes courtesy of 8 PCIe 4.0 lanes. In contrast, Intel supports 4 GB/s of throughput over its DMI link through the PCIe 3.0 interface. TRX40's extra bandwidth will benefit multiple devices that hang off the TRX40 chipset, like large PCIe SSD arrays.

Threadripper 3000 also offers up to 133GB/s of concurrent bandwidth to connected devices, while Intel only offers 52 GB/s, which is another significant advantage. 

MSI, Gigabyte, ASUS and ASRock have all announced leading-edge TRX40 motherboards, and we expect more to follow in the coming months. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Threadripper 3970X Boost Frequency

Given the impressively high boost clock specifications for these core-heavy chips, and AMD's continuing optimization of its boost implementation, we ran a few tests to measure the chip's ability to hit its rated speeds. 

Generally, stock coolers that worked with the previous-gen Threadripper models should suffice for most users, but beefier coolers can unlock more performance. AMD ships all Threadripper CPUs with an Asetek bracket that provides partial coverage of the massive heat spreader using supported closed-loop liquid coolers. According to AMD, this partial coverage is fine for stock operation, but we prefer full-coverage coolers. We tapped the Enermax Liqtech 360 TR4 II for cooling the processor at stock settings. AMD also collaborated with Cooler Master to develop the Wraith Ripper heat sink/fan combo for its Socket sTR4 interface. It's sold separately, though. We also used a beefy custom watercooling loop with two 360mm radiators for overclocking testing.

You'll need to install AMD's chipset drivers to enable the UEFI CPPC2 interface, which allows the operating system to prioritize scheduling threads into the fastest cores of the processor first. The chipset drivers also automatically install the required AMD-designed power plans. You'll also need the Windows May 2019 Update (or later) to enable the topology-aware scheduling implementation. Also, be aware that silicon quality, your motherboard, and firmware all play a role in the efficacy of AMD's unique boosting implementation, so your results may vary. 

With those prerequisites in place, we turned to our now-standard test that we've used in our series of Ryzen 3000 boost clock tests. We begin by recording the frequencies of each core during a series of commonly-used tests that should expose the peak frequencies. The first two tests are LAME and Cinebench in single-core test mode. These programs only execute on one core of the processor, which typically allows the chip to reach its peak boost frequency within its power, current, and thermal envelope. We also used tests with intermittent "bursty" workloads: PCMark 10, Geekbench, and VRMark run in rapid succession after the first two tests. 

With 32 cores hammering away, the per-core frequency recordings create unintelligible charts, so the album above only includes the maximum and minimum frequencies recorded during each 1-second measurement interval (100ms sampling). That means these measurements could come from any one core, but it makes the charts easier to digest. We've also plotted chip temperature on the right axis (the dark red line).

At stock settings with the AIO cooler, we reached 4.55 GHz frequently. You'll notice these boost frequencies are short in duration, while Intel's processors tend to reach their boost clock and stay there for longer periods of time (contingent upon BIOS settings). We engaged the auto-overclocking PBO feature with the Enermax AIO and recorded 4.575 GHz sporadically and received slightly more frequent boosts, but this varies on a run-to-run basis. 

Finally, we topped the processor with the custom loop and measured frequencies with PBO engaged. We recorded a few maximum boosts of 4.650 GHz, but we also recorded frequent boosts of 4.6 GHz across several test runs. Overall this is the most promising boost behavior we've seen with the Zen 2-based processors, and it appears that AMD's long-touted but never-realized promise of potential boosts beyond the rated peak frequency through its PBO algorithms may be realized, at least with our chip. We've flagged this for further investigation. 

We also examined the logging output and noticed that the boost activity frequently occurred in active cores. That's an improvement over past Ryzen firmwares that incorrectly boosted inactive cores, and signals that the improvements we've already seen on the X570 platform carry over to TRX40. 

Threadripper 3970X and 3960X Overclocking

AMD's Ryzen 3000 processors have drastically improved single-threaded performance, but you'll lose that benefit if you manually overclock. That's because the 7nm chips can't be manually overclocked on all cores to reach the same frequency as the single-core boost frequency. In fact, the all-core overclock ceiling is often 200 to 300 MHz lower than the single-core boost speeds, which is likely due to AMD’s new binning strategy that finds the Ryzen 3000 chips with a mix of both faster and slower cores. 

Given the already-prodigious power draw of these 280W TDP chips, temperatures are a concern for manual overclocking, though the solder thermal interface material (sTIM) between the heatspreader and dies does help thermal dissipation. Brute-force manual overclocking may not be the best path forward with conventional cooling, but newer BIOS revisions that support per-CCX overclocking have surfaced, which opens up a new pathway for more fine-grained optimizations. 

We turned to AMD's auto-overclocking Precision Boost Overdrive feature for our battery of tests. This auto-overclocking algorithm preserves the benefits of the single core boost, as seen in our boost testing above, while speeding up threaded workloads. We paired our PBO-enabled configurations with our custom watercooling loop and a Phanteks full-coverage Glacier C399A TR4 wateblock, enabling the utmost performance possible with our available cooling solutions. As with all Zen 2-based chips, PBO performance will vary based upon your cooling solution, motherboard, and firmware. 

Threadripper 3970X and 3960X Power Consumption

As we expect from processors with such high core counts, peak power measurements for the Threadripper 3000 processors outweigh all but the overclocked processors, but we do notice a substantial power reduction compared to the previous-gen Threadripper models. Given the level of performance, these power consumption figures are more than acceptable, and a quick look at the HandBrake efficiency metrics, which quantify the number of renders you can accomplish (given our workload) per day per watt of power consumed, reveals the processors offer a competitive efficiency profile. 

Meanwhile, the Core i9-10980XE is also very power efficient, notching impressive gains over the previous-gen Core i9-9980XE. The Threadripper processors are competitive, which is surprising given that they consist of four 7nm compute die and one 12nm I/O die, while the 10980XE uses a single monolithic die. AMD ties these five die together with the Infinity Fabric, but has obviously tuned the fabric to be incredibly power efficient. Notably, the Ryzen 9 3950X takes the crown as the most efficient processor in our test pool. 

Test Setup

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Test Notes

All AMD entries with "PBO" indicate an auto-overclocked configuration paired with with DDR4-3600. Intel's overclocked configurations also use DDR4-3600. 

AMD designed the Threadripper 3970X and 3960X for prosumer-class applications, but it still comes with a Game Mode preset in the Ryzen Master software that disables three of the four available dies (1/4). This feature debuted with the first-gen Threadripper processors to improve gaming performance and ensure compatibility with some games. 

AMD says this feature largely isn't needed anymore, although there are a few titles that aren't compatible with the copious helping of threads. Far Cry 5 notoriously struggles with high core counts, and we also noticed abnormally low performance and outwardly rough gameplay in Dawn of War: Warhammer. As such, we tested those titles in Game Mode with the 3970X and 3960X, but we tested the remainder of the games in the standard Creator Mode (all cores/threads active). 

The second-gen Threadripper models still suffer from the same odd performance in games if we leave all cores active, so we tested the Ryzen Threadripper 2990WX and 2770WX in game mode for all gaming tests.  

Given Threadripper 3000's high-priced nature, we fully expect these processors to be paired with high-resolution QHD (and beyond) displays. However, in keeping with our standard practice, we test at the FHD resolution to eliminate graphics-imposed bottlenecks. Be aware: These deltas will shrink at higher gaming resolutions. 

VRMark and 3DMark

The 3DMark DX12 and DX11 tests measure the amount of raw horsepower exposed by the processor to game engines, but most game engines don't scale as linearly with additional compute resources. 

The DX12 tests expose the huge step forward with the Threadripper 3000 as the chips easily outpace their predecessors. However, we can see this workload doesn't scale as linearly as we'd like with extra cores: The 24-core 48-thread 3960X outpaces the 32C/64T 3970X at both stock and overclocked settings, and easily outpaces the stock Cascade Lake-X Core i9-10980XE. However, Intel's new flagship turns the tables after overclocking.

The DX11 tests also don't scale as well with additional cores, though we do see the expected gains with overclocking. 

VRMark responds well to high per-core performance, but the Threadripper 3000 processors challenge Intel's finest HEDT chips: They both beat the HEDT competitors at stock settings. 

Intel's W-3175X takes the overall lead after we dial up the clocks, but bear in mind this is an exotic $3,000 chip that requires extremely extravagant accommodations.

Civilization VI AI and Stockfish

The Civilization VI AI test measures AI performance in a turn-based strategy game and is heavily influenced by high clock rates and instruction per cycle (IPC) throughput.

The top of these charts used to be Intel-only territory, but AMD has made amazing gains in per-core performance (a mixture of IPC and frequency) with the Zen 2 microarchitecture. Despite its hefty core counts, the Threadripper 3970X features the same boost clock as the 3960X, but it's premium silicon might be able to attain those boosts and stay at those heightened speeds for longer periods. Here we can see the 3970X beat the 3960X, if only by the slimmest of margins, and also experience additional uplift from the auto-overclocking feature. Intel's HEDT chips trail at stock settings, but once again take the lead after tuning. 

The open-source Stockfish AI chess engine is the polar opposite of the Civilization VI engine. This engine is designed specifically for many-core chips and scales well up to 512 cores, which is music to Threadripper's ears. At stock settings, the 3970X leads convincingly while the overclocked W-3175X struggles to keep pace. The 3960X once again doesn't see much uplift from overclocking, but it effectively ties the W-3175X at stock settings. 

Keep your eyes on the previous-gen Threadripper models as you flip through the charts. AMD's explosive gen-on-gen performance improvement, borne of a new architecture and manufacturing process, is impressive. 

Meanwhile in this test, the Cascade Lake-X i9-10980XE either ties or falls behind its predecessor, the -9980XE, at stock settings. 

Ashes of the Singularity: Escalation

Ashes of the Singularity: Escalation responds well to extra cores and threads, which benefits the Ryzen lineup. Intel's -10980XE suffers from a erratic frame latency during our test, and we tested multiple times and reinstalled the game/drivers in an attempt to rectify the issue. However, the condition is repeatable and carries over to the overclocked configuration, too. As we can see, this results in a lower 99th percentile frame rate, but that same trend applies to the W-3175X and the -9980XE. We theorize this stems from Intel's mesh architecture, present only on Intel's HEDT and data center processors, which can negatively impact performance with unoptimized software. It's also possible the issue is exacerbated by an early firmware revision for the refreshed X299 platform, or a lack of driver/game engine tuning.

Overclocking helps, but the -10980XE at 4.8 GHz trails the previous-gen -9980XE at 4.6 GHz. The overclocked W-3175X blasts to the top of the chart but its 99th percentile frame rates trail the 3970X. 

Meanwhile, the Threadripper processors are a solid generational step forward. As you can see at the bottom of the chart, the second-gen Threadripper chips aren't the best solution for gaming due to the eccentricities of their multi-die design. In contrast, Threadripper 3000 beats the stock Intel processors. 

Civilization VI Graphics Test

The Civilization VI graphics test finds the stock Ryzen 9 3950X delivering excellent performance given its price point. That reminds us that these HEDT processors aren't the best fit for gamers – most enthusiasts are better served by mid-range and high-end mainstream chips.

Intel's overclocking advantage comes into play once again, with the Core i9-10980XE taking a convincing lead. The stock Threadripper 3000 chips continue to impress, and the auto-overclocking PBO feature ekes out a few extra fps. 

Dawn of War III

It isn't surprising to see the overclocked Intel HEDT chips take the top of the Dawn of War III chart due to their per-core performance advantage. Again we see Threadripper 3000's big leap forward over the previous-gen models, but the gains from overclocking are muted. 

Intel's HEDT chips can suffer from their unique architecture in some game titles, but that same trend applies to the Threadripper processors, too. AMD's first- and second-gen Threadripper models were known for their erratic performance that stems from unoptimized game engines, necessitating a game mode that effectively removes cores and threads to boost performance. That trend doesn't plague Threadripper 3000 models as much, but there are exceptions. As a result, we tested the 3970X and 3960X in game mode for this title. You can see the difference in performance in the table below. 

Far Cry 5

Intel's -9900K leads the pack, and overclocking would open an even larger divide. The -10980XE is also particularly impressive after overclocking, but the 3970X is faster at stock settings.

As you can see in the table below, Far Cry 5's unoptimized code incurs a big performance penalty when all cores and threads are exposed to the operating system (creator mode), so we also tested this title in game mode. 

It's noteworthy that similar adjustments might also benefit Intel's high core-count processors, but the company hasn't made an easy-to-use tool to reduce core counts. AMD has instituted the ability to switch between game mode and creator mode in its Ryzen Master software, but it requires a reboot, so we still think it is an inconvenient band-aid.

Final Fantasy XV

We run this test with the standard quality preset to sidestep the impact of a bug that causes the game engine to render off-screen objects with the higher-resolution setting.

Intel's HEDT chips flex their gaming muscle when the game engine cooperates. Here the chips take the lead across the board at both stock and overclocked settings, pushing us close to a graphics-imposed bottleneck. 

Grand Theft Auto V

Grand Theft Auto V continues to be popular six long years after its release. This title favors Intel architectures and, more generally, multi-core designs with high clock rates. Intel's chips lead across the board in this title, and we spot a few significant outliers from both Threadripper 3000's and the Intel -10980XE that manifested as hitching during the benchmark sequence. 

Hitman 2

The Threadripper 3000 processors are competitive against Intel's processors at stock settings, but don't experience as much explosive uplift from overclocking. The Core i9-10980XE is impressive after tuning, nearly matching the W-3175X. 

Project Cars 2

Project Cars 2 is optimized for threading, but high clock rates pay off. As expected, that results in a win for Intel's overclocked processors. The 3970X leads the -10980XE at stock settings, while the 3960X is competitive. The Core i9-9900K is also impressive, and overclocking it would hand it the unequivocal win in this title. 

World of Tanks enCore 

AMD has addressed the lions share of the erratic gaming performance we observed with previous-gen models, but Intel's chips are beastly overclockers. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Test Notes

We ran an extra series of tests to reflect performance in workstation-class workloads. Some of these applications make an appearance in our standard test suite on the following page, but those test configurations and benchmarks are focused on a typical desktop-class environment. In contrast, these tests are configured to stress the systems with workstation-class workloads. 

With the exception of the W-3175X system, we loaded down our test platforms with 64GB of DDR4 memory spread across four modules to accommodate the expanded memory capacity required for several of these workstation-focused tasks. Due to the W-3175X's six channel memory controller and our limited stock of high-capacity DIMMs, we used six 8GB DIMMs for a total capacity of 48GB. All systems were tested at the vendor-specified supported memory data transfer rates for their respective stock configurations, and DDR4-3600 for the overclocked settings.  

We also conducted the tests on this page with a PCIe 4.0 Gigabyte Aorus SSD for all of the test systems, including the Intel platforms that are limited to PCIe 3.0 throughput. This will enable additional platform-level performance gains from the increased throughput of the faster interface supported by AMD's processors. 

All Threadripper processors are tested in creator mode, meaning the full heft of their prodigious number of cores and threads are in action. Due to the limited time available during the NDA test period, we only tested current-gen processors in overclocked configurations. 

Puget Systems Benchmarks

Puget Systems is a boutique vendor that caters to professional users with custom-designed systems targeted at specific workloads. The company has developed a series of acclaimed benchmarks for Adobe software, which you can find here. We use several of the benchmarks for our first round of workstation testing, followed by SPECworkstation 3 benchmarks.

Adobe After Effects CC Render Node Benchmark

The After Effects render node benchmark leverages the in-built aerender application that splits the render engine across multiple threads to maximize CPU and GPU performance. This test is memory-intensive, so capacity and throughput are important and can be a limiting factor.

The Threadripper processors deliver outstanding performance in this test, which isn't surprising given their hefty core counts. Yet again, Intel's processors benefit handsomely from overclocking, but it's noteworthy that most professional users will not spend the time tuning the system to extract the extra performance. Instead, stability at stock settings is typically the most prized attribute.

Adobe Premier Pro CC Benchmark

This benchmark measures live playback and export performance with several codecs at 4K and 8K resolutions. It also incorporates 'Heavy GPU' and 'Heavy CPU' effects that stress the system beyond a typical workload. Storage throughput also heavily impacts the score. 

Feeding the Threadripper processors with the throughput of PCIe 4.0 certainly helps, but sheer brute computational force appears to be the name of the game: The overclocked Xeon W-3175X challenges the 3970X even though it only has access to the PCIe 3.0 interface. However, the Threadripper processors enjoy a comfortable lead at stock settings. 

Aside from the second-gen Threadripper models, all of the chips sustain acceptable playback performance, but the real differentiation stems from the Export phase of the test. 

Adobe Photoshop CC Benchmark

The Photoshop benchmark measures performance in a diverse range of tasks, measuring the amount of time taken to complete general tasks, create panoramas (photomerge results) and apply filters. 

Once again the Threadripper processors prove their mettle with leading performance, but notice the Ryzen 9 3950X: That processor is significantly cheaper at $749 and drops into mainstream motherboards, which equates to lower overall platform pricing. It also only has access to a single dual-channel memory controller, yet manages to trade blows with Threadripper 3000's two dual-channel controllers. Given its pricing, and the performance you'll see throughout these workstation tests, the Ryzen 9 3950X steals the show.  

SPECworkstation 3 Benchmarks

The SPECworkstation 3 benchmark suite is designed to measure workstation performance in professional applications. The full suite consists of more than 30 applications split among seven categories, but we've winnowed down the list to tests that largely focus specifically on CPU performance. We haven't submitted these benchmarks to the SPEC organization, so be aware these are not official benchmarks. 

Media and Entertainment

We run the new Blender Benchmark beta in our regular suite of tests on the following page, but different types of render jobs can stress processors in unique ways. Here we can see a breakout of several industry-standard benchmark renders that largely favor Threadripper 3000's architecture. Again, the Ryzen 9 3950X is incredibly potent in some of these workloads, but Threadripper dominates at stock settings, while Intel leverages its overclocking headroom to squeeze out some additional gains. 

Threadripper steals the show in the Handbrake tests, too, and you'll notice a massive generational leap over the second-gen models. That is borne of AMD's revamped AVX engine. 

In the LuxRender CPU tests, the Xeon W-3175X is competitive with the 3970X at stock settings, and takes the crown after overclocking. Meanwhile, the Core i9-9980XE challenges the 3960X, but it takes quite a bit of tuning to get there. 

More workloads are leveraging the massive computational power of GPUs to accelerate key portions of parallelized workloads. The LuxRender GPU test shows us that all of the processors in our test pool offer similar performance with the task offloaded to the GPU (the slim variances in run-to-run performance are expected). 

NAMD and Rodinia LifeSciences

NAMD is a parallel molecular dynamics code designed to scale well with additional compute resources, and here we can see performance scale well across Threadripper's cores and threads through several permutations of the code. 

SPECworkstation 3's Rodinia LifeSciences benchmark steps through four tests that include medical imaging, particle movements in a 3D space, a thermal simulation, and image-enhancing programs. The Hotspot thermal simulation stands out as another big step forward for AMD's Threadripper chips. 

Product Development and Energy

The earth’s subsurface structure can be determined via seismic processing. One of the four basic steps in this process is the Kirchhoff Migration, which is used to generate an image based on the available data using mathematical operations. Threadripper 3000 takes a pronounced win here, shedding some of the comparatively lackluster performance of the previous-gen models.  

Calculix is based on the finite element method for three-dimensional structural computations. This benchmark performs well on the Intel processors, with the -10980XE being particularly impressive. Threadripper doesn't benefit much from tuning in this workload, but it is competitive. Again, the Ryzen 9 3950X pops into the conversation with surprisingly strong performance. 

SRMP algorithms are used for discrete energy minimization. AMD processors have traditionally struggled with these tests, suggesting the benchmark might be latency-sensitive, but Threadripper 3000 reverses that trend in astonishing fashion. You'll notice the stock Ryzen 9 3950X is faster than its own overclocked configuration. We'll have to rerun that test to verify the metric, but this fast-running benchmark seems to have more variation than others, meaning this could be an outlier. 

Financial and General Workloads

The Monte Carlo simulation is used to project risk and uncertainty in financial forecasting models. Here we see third-gen Threadripper post excellent performance metrics, with the previous-gen 32C/64T model matching the stock 24C/48T 3960X. Threadripper 3000 makes a clean sweep of the various financial models. 

The Python benchmark conducts a series of math operations, including numpy and scipy math libraries, with Python 3.6. This test also includes multithreaded matrix tests that would obviously benefit from more cores, provided the software can utilize the host processing resources correctly. Naturally, the multithreaded matrix workload favors Threadripper 3000, but the Intel processors dominate the numpy and scipy tests. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Test Notes

All systems in this section of tests use our standard test bench setup listed on the second page of the article. 

Rendering

As we seen on the preceding page, rendering is the perfect use case for Threadripper's generous helping of cores and threads. As we've seen with other Zen 2-based processors, Threadripper dominates in threaded workloads, and the improvements in single-core performance is impressive. Threadripper leads the single-core Cinebench benchmarks, but the stock Core i9-10980XE crops up in the single-core POV-Ray benchmark with leading performance due to its excellent performance with AVX workloads.

Intel's overclocked Xeon W-3175X takes the win in Corona and V-ray, but we expect that level of performance given it's $3,000 price tag and $1,500 motherboard. At stock settings, however, it often grapples with the $1,399 Threadripper 3960X, highlighting the massive pricing disparity. 

Encoding

Intel's Core i9-10980XE excels at single-threaded workloads due to its aggressive 4.6 GHz boost clock, so at settings it takes the lead over the Threadripper processors in the LAME benchmark. 

The Threadripper processors offer dominating performance at stock settings in the HandBrake x264 test, and in the AVX-heavy x265 version of that same benchmark. Flipping through to the SVT-AV1 encoder, which is heavily threaded, paints a similar picture. Yet again, the Threadripper 2990WX shows its bipolar nature as it lags in these tests due to its reduced memory throughput, highlighting AMD's solid performance gains with the same number of cores and threads.

Compression, Decompression, Encryption, AVX

The 7zip and Zlib compression/decompression benchmarks rely heavily upon threading and work directly from system memory, thus avoiding the traditional storage bottleneck in these types of tasks. 

The first-gen Threadripper processors are notorious for an unexplained deficiency in threaded 7zip compression workloads that find them trailing even the eight-core Core i9-9900K, but third-gen Threadripper marks a tremendous step forward in compression workloads. Threadripper leads the stock processors during these tests without storage throughput restrictions, but you also have access to the PCIe 4.0 interface with AMD's TRX40 platform. That extra throughput means these tremendous gains will largely transfer over to real-world application performance.

The heavily-threaded y-cruncher benchmark, which computes pi using the demanding AVX instruction set, is well-suited for the Xeon W-3175X and it's superior performance with AVX workloads. At stock settings the 28-core chip beats the 32-core Threadripper 3970X, and overclocking opens a larger gap. 

Office and Productivity

Microsoft's office suite runs via PCMark 10's new application test and uses real Microsoft Office applications. It seems like an odd fit to test these fire-breathing processors in such mundane tasks, but Office is ubiquitous. As we've come to expect, third-gen Threadripper is much more impressive in these workloads than its predecessor, but Intel remains very competitive in some of the subtests, particularly after overclocking. 

The application start-up metric measures load time snappiness in word processors, GIMP, and Web browsers. Other platform-level considerations affect this test as well, including the storage subsystem. The Threadripper 3970X falls to the middle of the pack, while Intel's chips leverage their frequency/overclocking advantage to take the lead. 

The timed-LLVM compilation workload is more well-suited to the 3970X, and overclocking offers a small performance boost. 

Web Browser

Browsers tend to be impacted more by the recent security mitigations than other types of applications, so Intel has generally taken a haircut in these benchmarks of fully-patched systems. Unsurprisingly, the Ryzen 9 3950X and Core i9-9900K are pretty agile in these workloads, but the Core i9-10980XE in stock trim is plenty snappy, too. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Intel's seemingly-endless delay in transitioning to the 10nm node for the desktop, not to mention a new architecture beyond Skylake, has left the industry ripe for disruption. As a result, AMD's new Threadripper 3000 processors march into the upper segment of the HEDT market uncontested. 

Intel does have its workstation-focused Xeon W series, but the company's expensive pricing model for those chips, not to mention the supporting platforms, is now under serious attack. It wouldn't be surprising to see AMD's forthcoming 64-core 128-thread Threadripper 3990X debut at lower pricing than Intel's flagship $4,449 Xeon W-3275 that only comes with 28 cores. That means you can expect deep price cuts from Intel in that segment, too, to match the price cuts it made as it moved forward to its Cascade Lake-X processors. 

Threadripper 3970X and 3960X mark a significant step forward in nearly every way, either reducing or eliminating many of the vagaries of the first- and second-gen Threadripper processors. Unlike the previous-gen chips, there were no errant results that would give us pause in making a universal recommendation.

The Zen 2 architecture provides a tangible bump in instruction per cycle (IPC) throughput. Pairing that advance with AMD's clever combination of software, firmware, and hardware to target the fastest cores with lightly-threaded tasks yields a big advance in overall per-core performance. AMD spreads that performance out over multiple cores with a more uniform design that eliminates many of the odd performance issues that prevented Threadripper from enjoying broader uptake in the professional segment. Not to mention that the new design allows AMD to utilize more efficient manufacturing techniques that enable paradigm-shifting changes to the pricing landscape. 

AMD has bumped up pricing for the Threadripper 3000 processors, and a quality TRX40 motherboard isn't going to come cheap. You'll also need a capable cooler to dissipate heat from the beefy 32-core chips to extract the most performance from the silicon, along with a quality power supply. These are expected accommodations for a processor of this class, and given the performance we saw throughout our test suite, AMD's per-core asking price is very competitive. 

The Threadripper 3970X and 3960X delivered devastating threaded performance in their respective price ranges, often trouncing Intel's most exotic silicon. Intel's Xeon W-3175X is ill-suited to take on the comparatively power-sipping Threadripper processors on a power efficiency basis, not to mention pricing. Just for comparison's sake – the overclocked W-3175X pulled 768 watts under load, while the overclocked Threadripper 3970X peaked at 356 watts while often providing more performance in threaded workloads. That math is easy. 

Threadripper 3000 also brings a solid gain on the single-threaded performance front, too. As a general observation, both Threadripper 3970X and 3960X offered surprising snappiness in general desktop use, which is a first for any Threadripper processor we've tested. 

Finally, AMD's forward-thinking adoption of the PCIe 4.0 interface is another attraction that will help win over the semi-professional crowd. While the faster interface isn't as useful on the mainstream desktop, the ability to stack up throughput-craving devices behind the chipset without the radical throughput restrictions we see with Intel's DMI is a win. 

Intel has cut the pricing of its Cascade Lake-X processors in anticipation of Threadripper 3000, which promises to shake up the value proposition at lower price ranges. We'll analyze that in detail in our coming review. 

For now, the highest tiers of the HEDT market belong to AMD's Threadripper 3000 processors. If you're looking for a chip that is incredibly powerful for brute-force parallelized workloads, but still agile enough for some entertainment, the Threadripper 3970X is the perfect solution. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Intel's launch of the 10th-generation Cascade Lake-X processors marks yet another iteration of its 14nm process, serving as a refresh of its Skylake-X refresh processors. The Core i9-10980XE slots in as the flagship with a steep gen-on-gen price cut that reduces pricing from $1,999 to a 'mere' $979 for 18 cores and 36 threads. 

Intel's price cuts come as a byproduct of AMD's third-gen Ryzen and Threadripper processors, with the former bringing HEDT-class levels of performance to mainstream 400- and 500-series motherboards, while the latter lineup is so powerful that Intel, for the first time in recent history, doesn't even have a response. 

AMD is moving forward at breakneck speed with the 7nm process and Zen 2 architecture, but Intel remains mired on the 14nm process. Due to the limitations of the 14nm manufacturing process, Intel simply doesn't have room to add more cores, let alone deal with the increased heat, within the same package. Instead, Cascade Lake-X offers a few incremental bumps in frequency, exposes four more PCIe 3.0 lanes, and comes infused a few more hardware mitigations to deal with Intel's expanding list of security vulnerabilities. 

That means Intel's price cuts serve as the most impressive feature in its new lineup, allowing the company to seek shelter from AMD's Threadripper in what is now the low end of the HEDT market. 

But even though it isn't a direct competitor, Intel can't hide from the 16-core 32-thread Ryzen 9 3950X. The -10980XE is thoroughly unimpressive from a new technology standpoint, but Intel's incessant iterations of the 14nm process have yielded higher overclocking potential, lower power consumption, and incremental improvements in memory support that deliver decent bumps in performance. However, AMD's Ryzen 9 3950X still takes the shine off the -10980XE for users that don't need quad-channel memory. AMD has also left its second-gen Threadripper processors on the market as the 'value' HEDT alternative, but they aren't nearly as competitive as third-gen Ryzen. 

Intel Cascade Lake-X Core i9-10980XE Pricing and Specifications

The 9th-gen processors have an average price-per-core of $103, but Intel reduced to ~$57 per core for 10th-gen Cascade Lake-X that is based on the same Skylake microarchitecture. The Core i9-10980XE slots in with 18 cores at 3.0GHz and a TDP of 165W. It also has a slightly improved Turbo Boost 2.0 frequency of 4.6GHz and 4.8GHz with Turbo Boost 3.0. 

The 10-core Core i9-10900XE sits at the bottom with slightly improved base and boost frequencies and will cost $590, which is only $30 more than the expected pricing of the Core i9-9900KS for the mainstream desktop. Intel says that it made this drastic price cut on the low end to make it easier for customers to step up to the more capable platform with its expanded feature set.

Intel doesn't have a 16-core model to deal with AMD's much-anticipated Ryzen 9 3950X. That might mean Intel has something in store for a later date, as we doubt the company will leave the 3950X uncontested.

We expect Intel to address the higher core count Threadripper with a refreshed Xeon W lineup, but as we can see from the $4,485 W-3275 and $3,349 W-3265, these chips come with eye-watering pricing that simply isn't sane in a post-Threadripper 3000 world. We expect Intel will slash the pricing on these chips, too, but they drop into server-derived sockets, so platform pricing will continue to be a problem for Intel. They also aren't overclockable – yet. 

Intel also has its Xeon W-3175X, the lone overclockable processor on the LGA 3647 platform, for $3,000. This 28-core 56-thread processor drops into exotic motherboards that carry heart-stopping price tags, so it really isn't a direct competitor to Threadripper 3000, either.

Intel bumped up clock speeds across the entire range and added a new boosting feature that targets up to four cores with its Turbo Boost 3.0 technology, an expansion of the feature that allows it to hit two more cores with lightly-threaded applications. However, the frequency for the second two fastest cores drops 100 MHz from the listed Turbo Boost Max 3.0 frequency.

Intel increased memory support to DDR4-2933 and a maximum capacity to 256GB. Intel also brought support for DL Boost, Intel’s new AI-focused instructions that double or triple the performance of FP16 or INT8 AVX-512 vector code, to the HEDT space.

The Cascade Lake-X chips come with 48 PCIe lanes, an increase of four extra lanes compared to the previous-gen models. Intel exposed four additional lanes from the HCC die to the socket, and while these chips are compatible with existing X299 boards, you'll lose those extra four lanes unless you upgrade to one of the new X-series motherboards that also debuted with the new chips. These motherboards are a decent step forward with enhanced secondary features, but Intel still remains on PCIe 3.0, a disadvantage compared to AMD's support for the PCIe 4.0 interface that offers twice the I/O bandwidth. 

For HEDT, AMD also still holds the lead in sheer PCIe lane counts at 60. It also holds the overall core count title with 32 for its Threadripper 2970WX that AMD says competes as a 'value HEDT' processor for this price range, while the Ryzen 9 3950X has a disadvantage with 24 PCIe lanes and a dual-channel memory controller. However, the 3950X provides an amazing amount of performance given its faster PCIe interface and is incredibly efficient given its available memory throughput.  

A Whole Lotta Turbo

However, Intel also increased multi-core turbo ratios, too. Here we can see Intel's three-tiered Turbo system, with differently frequencies for SSE, AVX2, and AVX-512 instructions. The -10980XE features heightened boost speeds for standard SSE and AVX2 across up to 12 cores, while AVX-512 gets a boost for up to eight cores. 

Intel managed to bump up these multi-core turbo ratios, which equate to more performance in real-world workloads, while delivering impressive power reductions that we'll cover on the next page. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Core i9-10980XE Boost Frequency

We ran a few tests to measure the Core i9-10980XE's ability to hit its rated boost speeds. Intel targets the fastest cores, in this case four of them, with lightly-threaded workloads with its Turbo Boost Max 3.0 feature. However, the frequency for the second two fastest cores drops 100 MHz from the listed Turbo Boost Max 3.0 frequency. Intel pioneered this technique for the desktop, but AMD has also adopted the tactic to extract more performance from its third-gen Ryzen chips.

We begin by recording the frequencies of each core during a series of commonly-used tests that should expose the peak frequencies. The first two tests are LAME and Cinebench in single-core test mode. These programs only execute on one core of the processor, which typically allows the chip to reach its peak boost frequency within its power, current, and thermal envelope. We also used tests with intermittent "bursty" workloads: PCMark 10, Geekbench, and VRMark run in rapid succession after the first two tests. 

The album above only includes the maximum and minimum frequencies recorded during each 1-second measurement interval (100ms sampling). That means these measurements could come from any one core, but it makes the charts easier to digest. We've also plotted chip temperature on the right axis (the dark red line).

As you can see, both chips can hit their rated turbo frequencies rather easily, and frequently, with the Core i9-10980XE offering a 300MHz improvement over the previous-gen -9980XE in single-threaded work. We also noticed the impact of the expanded four-core Turbo Boost Max 3.0 as more cores engage in boost activity during lightly-threaded workloads.  

Core i9-10980XE Overclocking

Overclocking still remains a key Intel advantage in the face of AMD's brutally-powerful Zen 2 processors. Intel's chips have far higher overclocking headroom and are realtively easy to tune manually, and deliver a much larger performance improvement than we see with AMD procesors. However, you'll need beefy cooling to wring out peak overclocking performance from the -10980XE, but it generally offers much higher overclocking headroom than the -9980XE.

We used our go-to Corsair H115i cooler for stock testing, but switched over to a custom loop with two 360mm radiators for overclocking. The -10980XE allows you to adjust bins for three types of instructions: IA/SSE, AVX2, and AVX-512.

We adjusted the -10980XE turbo ratios to 4.8, 4.0, and 3.3 GHz, respectively. We dialed these in as all-core overclock values, which means we do lose a bit of steam in lightly-threaded AVX-512 workloads that boost up to 3.8GHz on two cores, but you can also assign these values on a per-core basis to recoup those losses. As you can see, our -10980XE sample overclocked far beyond the limits we found with our -9980XE, but your mileage may vary. 

Overclocking was a straightforward affair: We bumped memory speeds up to DDR4-3600 and used a 2.1v VCCIN paired with a 1.2V vCore. We also bumped up the mesh multiplier to 32 and increased vMesh to 1.2V, which imparts a nice performance kicker due to better cache latencies. We also uncorked all of the power limits in the BIOS. Temperatures with our custom loop peaked at 80C during AVX2 workloads, but be aware that standard closed-loop coolers will be a limiting factor.

Even with relaxed settings, heat easily overwhelmed our H115i after small voltage increases, and that was with its fans cranking away at maximum speed. Just like the first- and second-gen Skylake-X chips, thermals limited our overclocking efforts before hitting the silicon's limits, despite solder-based TIM. Build your own custom loop if you plan on serious overclocking. Also, we advise forced air or water cooling on the power delivery subsystem. Invest in a PSU able to deliver at least 20A on the +12V rail. MSI’s BIOS warns that you need a power supply capable of providing up to 1000W through the eight-pin EPS cable; a beefy PSU is non-negotiable.

Our overclocking technique for AMD's 3000-series processors is drastically different. These chips offer improved single-threaded performance, but you'll lose that benefit if you manually overclock. That's because the 7nm chips can't be manually overclocked on all cores to reach the same frequency as the single-core boost frequency. In fact, the all-core overclock ceiling is often 200 to 300 MHz lower than the single-core boost speeds, which is likely due to AMD’s new binning strategy that finds the Ryzen 3000 chips with a mix of both faster and slower cores. 

We turned to AMD's auto-overclocking Precision Boost Overdrive feature for our battery of tests. This auto-overclocking algorithm preserves the benefits of the single core boost, as seen in our boost testing above, while speeding up threaded workloads. We paired our PBO-enabled configurations with our custom watercooling loop and a Phanteks full-coverage Glacier C399A TR4 wateblock, enabling the utmost performance possible with our available cooling solutions. As with all Zen 2-based chips, PBO performance will vary based upon your cooling solution, motherboard, and firmware. 

Core i9-10980XE Power Consumption

Sometimes we encounter results that simply seem to good to be true, and that is exactly what transpired with the -10980XE's power measurements. We retested both the -10980XE and -9980XE on multiple motherboards to verify the power consumption deltas, and while they vary based on motherboard, there is no doubt the -10980XE is significantly more power efficient. We also used Intel's XTU to monitor both chips during stress testing and noticed that the -9980XE encountered current/EDP limit throttling much more frequently than the -10980XE during extended AVX workloads, indicating that motherboard vendors have a bit more budget to play with on the Cascade Lake-X chip. This could be due to refinements Intel has made to power thresholds, but most of the secret sauce behind these power reductions are hidden in the lines of BIOS code. 

In either case, the -9980XE draws ~20% more power than the -10980XE with both SSE and AVX2 instructions. We also measured peak AVX-512 power draw (not in the charts) at 149W for the -10980XE and 162W for the -9980XE, an 8% increase for the older chip. Normally we would expect these kind of power enhancements to stem from a newer, small process, but Intel has definitely mastered its 14nm, which you would expect after five long years and countless revisions. 

However, those large deltas in stress tests don't always carry over to all real world workloads. The -9980XE only drew 2W less than the -10980XE during the y-cruncher workload, though it is noteworthy that the latter was ~9% faster during the test, indicating that it is more efficient. Power efficiency gains will definitely vary based upon workload.

Flipping over to the AVX-enabled HandBrake tests finds the -10980XE drawing 16 fewer watts in the x265 workload, which has a heavier distribution of AVX instructions than the x264 test. In x264, the -10980Xe drew 14 fewer watts.

At idle, the -10980XE measured 26.5W, compared to the -9980XE's 28.8W. 

These improvements are impressive, but a glance at the Ryzen 9 3950X reveals the power efficiency of the new 7nm process, albeit paired with a 12nm I/O die. When we compare that chip to the -10980XE in the final two charts, which calculate power efficiency based on performance, the 3950X tops the chart for the Handbrake tests. The -10980XE also notches an impressive efficiency gain over the -9980XE in these real-world workloads, but the more we see of the 3950X, the more we like it. 

By virtue of its drastically lower price point, the -10980XE doesn't compete directly with the 3970X and 3960X, but it is more power efficient. There is some overhead for AMD's Threadripper chips due to a larger 12nm central I/O die and the Infinity Fabric that it uses to tie together the four chiplets, but they are even more power efficient than the eight-core -9900K with its monolithic die, which is impressive. Also, a glance at the 2990WX and 2970WX in the efficiency charts really highlights the transformational gen-on-gen efficiency improvement. 

Test Setup

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Test Notes

All AMD entries with "PBO" indicate an auto-overclocked configuration paired with with DDR4-3600. Intel's overclocked configurations also use DDR4-3600. Our Threadripper 2970WX sample doesn't respond well to AMD's auto-overclocking PBO feature, generating incessant BSOD's even after numerous hours of experimentation. That issue seems confined to that sample; our other Threadripper processors behave correctly with the feature enabled. As such, we can't provide benchmarks with an overclocked 2970WX, but the stock test results fall within our expectations.

We tested Far Cry 5 and Dawn of War in Game Mode with the 3970X and 3960X, but we tested the remainder of the games in the standard Creator Mode (all cores/threads active). We tested the Ryzen Threadripper 2990WX and 2770WX in game mode for all gaming tests. 

We expect these benchmark deltas to shrink with higher resolutions that are more typical of the class of machines with these chips, but that is just due to a graphics-imposed bottleneck. As such, we stick to the standard FHD resolution for testing. 

VRMark and 3DMark

Intel has two AMD competitors in these benchmarks: The Ryzen 9 3950X and Threadripper 2970WX, with the former delivering impressive gaming performance in games, while the latter clearly isn't as agile and has to run in game mode with a portion of its threaded horsepower left unused. This mode requires a reboot and has a varying impact on different games, so your mileage will vary. In either case, incessant reboots aren't a great selling point if you game frequently. AMD has largely addressed those concerns with third-gen Threadripper, but doesn't have a competing 3000-series chip in the -10980XE's price bracket. 

The 3DMark DX12 and DX11 tests measure the amount of raw horsepower exposed by the processor to game engines, but most game engines don't scale as linearly with additional compute resources. These tests reward the -10980XE's overclocking prowess with large leads after tuning, but the Ryzen 9 3905X leads at stock settings. We can also see the impact of the -10980XE's heightened mid-range turbo boosts in the thread-friendly DX12 benchmark as it surpasses the -9980XE by a decent margin. 

VRMark responds well to high per-core performance, so overclocking pushes the -10980XE to the top of the heap. The chip provides slightly more performance than its predecessor at stock settings, with the extra 300MHz of boost speed leading to a ~7fps advantage. Again, the 3950X outpaces the stock -10980XE.

Civilization VI AI and Stockfish

The Civilization VI AI test measures AI performance in a turn-based strategy game and is heavily influenced by high clock rates and instruction per cycle (IPC) throughput.

Here we see the 3950X edge out the -10980XE at stock settings, but in what will become a recurring theme in this set of benchmarks, Intel's flagship is impressive after overclocking. However, the 3950X isn't too shabby after overclocking, either. Intel's slim advantage after tuning comes at a $230 premium and requires a more robust cooling solution, so keep that in mind as you flip through the results. Meanwhile, the 2970WX isn't competitive, notching the lowest performance of the second-gen Threadripper lineup. 

The open-source Stockfish AI chess engine is the polar opposite of the Civilization VI engine. This engine is designed specifically for many-core chips and scales well up to 512 cores, which is music to third-gen Threadripper's ears. Those models blast to the top of the charts, but aren't relevant in this price class. The 3950X is impressive at stock settings, and the -10980XE trails the -9980XE slightly, indicating that it isn't going to be a universal step forward. They also tie in the Civilization AI test. However, the -10980XE does expose some additional overclocking headroom that yields big gains.  

AMD's explosive gen-on-gen Threadripper performance improvement, borne of a new architecture and manufacturing process, is impressive, but the 2970WX suffers at the hands of its distributed memory architecture. 

Ashes of the Singularity: Escalation

Ashes of the Singularity: Escalation responds well to extra cores and threads, which benefits the Ryzen lineup. Intel's -10980XE suffers from a erratic frame latency during our test, and we tested multiple times and reinstalled the game/drivers in an attempt to rectify the issue. However, the condition is repeatable and carries over to the overclocked configuration, too. As we can see, this results in a lower 99th percentile frame rate, but that same trend applies to the W-3175X and the -9980XE. We theorize this stems from Intel's mesh architecture, present only on Intel's HEDT and data center processors, which can negatively impact performance with unoptimized software. It's also possible the issue is exacerbated by an early firmware revision for the refreshed X299 platform, or a lack of driver/game engine tuning. We also notice the -10980XE trails the -9980XE in both stock and overclocked configurations. 

Overclocking helps, but the -10980XE at 4.8 GHz trails the previous-gen -9980XE at 4.6 GHz. Threadripper 2970WX trails the pack again. 

Civilization VI Graphics Test

The Civilization VI graphics test finds the stock Ryzen 9 3950X delivering excellent performance given its price point. That reminds us that these HEDT processors aren't the best fit for gamers – most enthusiasts are better served by mid-range and high-end mainstream chips.

Intel's overclocking advantage comes into play once again, with the Core i9-10980XE taking a convincing lead. The 2970WX simply isn't competitive here, even when we consider its slightly lower price point.

Dawn of War III

It isn't surprising to see the overclocked Intel HEDT chips take the top of the Dawn of War III chart due to their per-core performance advantage. Whatever issues plague the -10980XE in some titles aren't a factor here: Intel's new chip takes a big step forward over the -9980XE. 

Again, AMD's Ryzen 9 3950X is impressive here, but the -9900K is faster at a much lower price point, but you'll lose out on performance in threaded applications. 

Far Cry 5

Intel's -9900K leads the pack, and overclocking would open an even larger divide. The -10980XE is also particularly impressive after overclocking, but trails the 3950X at stock settings. After overclocking, the 2950X uncharacteristically experiences a big jump in performance with this title, but it largely profits from its overclocked memory.

Final Fantasy XV

We run this test with the standard quality preset to sidestep the impact of a bug that causes the game engine to render off-screen objects with the higher-resolution setting.

Intel's HEDT chips flex their gaming muscle when the game engine cooperates. Here the chips take the lead across the board at both stock and overclocked settings, pushing us close to a graphics-imposed bottleneck. The Core i9-10980XE also takes a decent step forward over the -9980XE at stock settings. 

The Ryzen 9 3950X trails substantially, and the 2970WX continues to be a non-factor in the gaming conversation. 

Grand Theft Auto V

Grand Theft Auto V continues to be popular six long years after its release. This title favors Intel architectures and, more generally, multi-core designs with high clock rates. Intel's chips lead across the board in this title, and we spot a few significant outliers from both Threadripper 3000's and the Intel -10980XE that manifested as hitching during the benchmark sequence. 

The Ryzen 9 3950X trails the stock -10980XE by 2.7 fps. Tuning the 3950X essentially yields a tie, but turning the dial to 4.8GHz on the -10980XE propels it into rarefied air.  

Hitman 2

Once again, overclocking the -10980XE enables chart-topping performance and the chip also delivers a nice bump over its predecessor at stock settings. The 3950X isn't as impressive in this title, and engaging the auto-overclocking PBO feature doesn't deliver much extra performance. 

Project Cars 2

Project Cars 2 is optimized for threading, but high clock rates pay off. As expected, that results in a win for Intel's overclocked processors. The stock -10890XE handily beats the 2970WX, but trails the 3950X. 

The Core i9-9900K is also impressive, and overclocking it would hand it the unequivocal win in this title. 

World of Tanks enCore 

The power of Intel overclocking in gaming is apparent again as the -10980XE separates itself from the rest of the test pool. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

SPEC Workstation and Adobe Performance Test Notes

We ran an extra series of tests to reflect performance in workstation-class workloads. Some of these applications make an appearance in our standard test suite on the following page, but those test configurations and benchmarks are focused on a typical desktop-class environment. In contrast, these tests are configured to stress the systems with workstation-class workloads. 

With the exception of the W-3175X system, we loaded down our test platforms with 64GB of DDR4 memory spread across four modules to accommodate the expanded memory capacity required for several of these workstation-focused tasks. Due to the W-3175X's six channel memory controller and our limited stock of high-capacity DIMMs, we used six 8GB DIMMs for a total capacity of 48GB. All systems were tested at the vendor-specified supported memory data transfer rates for their respective stock configurations, and DDR4-3600 for the overclocked settings. 

We also conducted the tests on this page with a PCIe 4.0 Gigabyte Aorus SSD for all of the test systems, including the Intel platforms that are limited to PCIe 3.0 throughput. This will enable additional platform-level performance gains from the increased throughput of the faster interface supported by AMD's processors. 

All Threadripper processors are tested in creator mode, meaning the full heft of their prodigious number of cores and threads are in action.

Puget Systems Benchmarks

Puget Systems is a boutique vendor that caters to professional users with custom-designed systems targeted at specific workloads. The company developed a series of acclaimed benchmarks for Adobe software, which you can find here. We use several of the benchmarks for our first round of workstation testing, followed by SPECworkstation 3 benchmarks.

Adobe After Effects CC Render Node Benchmark

The After Effects render node benchmark leverages the in-built aerender application that splits the render engine across multiple threads to maximize CPU and GPU performance. This test is memory-intensive, so capacity and throughput are important and can be a limiting factor.

Focusing on the price-competitive processors, the stock -10980XE leads the 2970WX and Ryzen 9 3950X, but the 3950X profits heavily from overclocking. Bumping up the voltage on the -10980XE opens up a large divide between it and the rest of the test pool.

Adobe Premier Pro CC Benchmark

This benchmark measures live playback and export performance with several codecs at 4K and 8K resolutions. It also incorporates 'Heavy GPU' and 'Heavy CPU' effects that stress the system beyond a typical workload. Storage throughput also heavily impacts the score. 

Feeding the third-gen AMD processors with the throughput of PCIe 4.0 certainly helps, but sheer brute computational force appears to be the name of the game. The Ryzen 9 3950X trails the -10980XE at stock settings, but overclocking again pushes the -10980XE into the lead.

Adobe Photoshop CC Benchmark

The Photoshop benchmark measures performance in a diverse range of tasks, measuring the amount of time taken to complete general tasks, create panoramas (photomerge results) and apply filters. 

The Threadripper processors prove their mettle with leading performance, but notice the Ryzen 9 3950X: That processor is significantly cheaper at $749 and drops into mainstream motherboards, which equates to lower overall platform pricing. It also only has access to a single dual-channel memory controller, yet manages to trade blows with Threadripper 3000's two dual-channel controllers. Given its pricing, and the performance you'll see throughout these workstation tests, the Ryzen 9 3950X steals the show.  

SPECworkstation 3 Benchmarks

The SPECworkstation 3 benchmark suite is designed to measure workstation performance in professional applications. The full suite consists of more than 30 applications split among seven categories, but we've winnowed down the list to tests that largely focus specifically on CPU performance. We haven't submitted these benchmarks to the SPEC organization, so be aware these are not official benchmarks. 

Media and Entertainment

We run the new Blender Benchmark beta in our regular suite of tests on the following page, but different types of render jobs can stress processors in unique ways. Here we can see a breakout of several industry-standard benchmark renders that largely favor the Zen 2 architecture. 

The Ryzen 9 3950X is incredibly potent in the Blender and Handbrake benchmarks, leading the -stock -10980XE in the majority of the tests. Meanwhile, the 2970WX isn't as competitive with its price bracket, especially when you consider the 3950X's lower platform costs. Overclocking flips the script in favor of Intel's Core i9-10980XE, but but that gain requires a more expensive platform and cooler, not to mention populating a quad-channel memory controller. However, the seconds measured in these tests can turn to hours for prolific users, so gearing the purchasing decision towards your workload is critical.

The LuxRender benchmarks favor the -10980XE at both stock and overclocked settings.

More workloads are leveraging the massive computational power of GPUs to accelerate key portions of parallelized workloads. The LuxRender GPU test shows us that all of the processors in our test pool offer similar performance with the task offloaded to the GPU (the slim variances in run-to-run performance are expected). 

NAMD and Rodinia LifeSciences

NAMD is a parallel molecular dynamics code designed to scale well with additional compute resources, and again the Ryzen 9 3950X pulls ahead of the -10980XE at stock settings, but trails after overclocking. 

SPECworkstation 3's Rodinia LifeSciences benchmark steps through four tests that include medical imaging, particle movements in a 3D space, a thermal simulation, and image-enhancing programs. The -10980XE inexplicably suffers in the heartwall simulation benchmark, but proves to be more competitive in the other Rodinia tests, indicating this may be another workload that doesn't 'mesh' well with Intel's mesh architecture.

Product Development and Energy

The earth’s subsurface structure can be determined via seismic processing. One of the four basic steps in this process is the Kirchhoff Migration, which is used to generate an image based on the available data using mathematical operations. The same trend emerges: The -10980XE trails the 3950X at stock settings, but benefits from tuning. 

Calculix is based on the finite element method for three-dimensional structural computations. This benchmark performs well on the Intel processors, with the -10980XE being particularly impressive. Again, the Ryzen 9 3950X pops delivers surprisingly strong performance, but doesn't benefit much from overclocking.  

SRMP algorithms are used for discrete energy minimization. The -10980XE wins this benchmark in convincing fashion, while the 2970WX suffers at the hands of its distributed memory architecture. 

Financial and General Workloads

The Monte Carlo simulation is used to project risk and uncertainty in financial forecasting models. Intel's finest trails the test pool at stock settings, once again falling to the mighty 3950X. It even trails the 2970WX in these tests.

The Python benchmark conducts a series of math operations, including numpy and scipy math libraries, with Python 3.6. This test also includes multithreaded matrix tests that would obviously benefit from more cores, provided the software can utilize the host processing resources correctly. Naturally, the multithreaded matrix workload favors Threadripper 3000, but the Intel processors dominate the numpy and scipy tests. This is due to their compilation in Intel's MKL library, but alternative libraries would improve AMD's standing.  

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Application Performance Test Notes

All systems in this section of tests use our standard test bench setup listed on the second page of the article. 

Rendering

Our rendering test suite finds Intel's processors in hostile territory: Multi-threaded rendering has become the domain of Zen-based architectures and their core-heavy designs. Threadripper 2970WX leverages its 24 cores and 48 threads to upset the stock -10980XE in several of the threaded rendering benchmarks, and tuning would improve its performance further. The Ryzen 9 3950X also impresses in the threaded LuxMark and POV-Ray tests.

The stock Core i9-10980XE is competitive in the single-core POV-Ray benchmark due to its strong performance in AVX workloads, but the 3950X takes the lead in both the single-threaded POV-Ray and Cinebench benchmarks.

Encoding

Intel's Core i9-10980XE excels at single-threaded workloads due to its aggressive 4.8 GHz boost clock, so it takes the lead over the AMD processors in the LAME benchmark at stock settings. 

The 3950X beats the stock -10980XE in the HandBrake x264 test, and in the AVX-heavy x265 version of that same benchmark. Flipping through to the SVT-AV1 encoder, which is heavily threaded, paints a similar picture. 

Compression, Decompression, Encryption, AVX

The 7zip and Zlib compression/decompression benchmarks rely heavily upon threading and work directly from system memory, thus avoiding the traditional storage bottleneck in these types of tasks. 

The first-gen Threadripper processors are notorious for an unexplained deficiency in threaded 7zip compression workloads that find them trailing even the eight-core Core i9-9900K, but third-gen Threadripper marks a tremendous step forward in compression workloads. Meanwhile, the 3950X doesn't fare quite as well in compression, but proves nimble with decompression as it easily beats the stock -10980XE. 

The 2970WX suffers during the single-threaded y-cruncher benchmark, which computes pi using AVX instructions, while the 3950X is far more competitive. That said, the -10980XE caps an all-Intel win. 

Office and Productivity

Microsoft's office suite runs via PCMark 10's new application test and uses real Microsoft Office applications. It seems like an odd fit to test these fire-breathing processors in such mundane tasks, but Office is ubiquitous. As we've come to expect, the overclocked -10980XE leads the tests, but the Ryzen 9 3950X challenges, and often beats, the stock Core i9-10980XE. 

The application start-up metric measures load time snappiness in word processors, GIMP, and Web browsers. Other platform-level considerations affect this test as well, including the storage subsystem.

Web Browser

Browsers tend to be impacted more by the recent security mitigations than other types of applications, so Intel has generally taken a haircut in these benchmarks of fully-patched systems. Unsurprisingly, the Ryzen 9 3950X and Core i9-9900K are pretty agile in these workloads, but the Core i9-10980XE in stock trim is plenty snappy, too. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Intel's tactic of slowly bumping up clock speeds and adding more features across its product stack, like Hyper-Threading, has proven to be a good-enough strategy to fend off AMD's increasing pressure with the first-gen Zen chips, but the arrival of Zen 2 and the 7nm process blow that approach out of the water. It's quite shocking to see Intel thoroughly unprepared to attack AMD's high end Threadripper parts, and we're not convinced that bringing the high core count Xeon W parts down to the standard HEDT segment would help.

Intel says it will have 10nm parts for the desktop soon, but we don't know where those products will land yet, and they certainly won't attack the HEDT market for at least another year, meaning the company has ceded the high end to AMD.

So what's left? Competing on price by dropping Cascade Lake-X pricing roughly 50% across the entire stack, thus dealing with AMD's lesser-equipped processors. That does improve Intel's value proposition, but AMD still looms large.

The refined 14nm process equates to faster clocks speeds, and thus performance, at lower overall power consumption. The Core i9-10980XE also has much higher overclocking headroom than its predecessor, but Intel's textbook incrementalism is no longer enough to fend off AMD in the 7nm Zen 2 era. 

The Core i9-10980XE has two primary competitors: the Ryzen 9 3950X and the Threadripper 2970WX that AMD has left on the market to address this price bracket. Due to the 2970WX's bipolar performance trends, with excellent performance in some threaded workload but flaccid performance in others, it's hard to recommend that chip as a general-purpose alternative. There are a slim cross-section of workloads where the 2970WX makes sense, but you'll suffer too many trade-offs in consistent performance across all workloads. 

The Ryzen 9 3950X looms large as the unlikely competitor for the -10980XE. The chip lacks a quad-channel memory controller and doesn't come with as many PCIe lanes, but its PCIe 4.0 lanes deliver twice the throughput of Intel's PCIe 3.0-equipped -10980XE. Given the Ryzen 9 3950X's impressive performance in many mainstream tasks, including threaded workloads, and its lower overall chip and platform pricing, it is a viable alternative to Intel's Core i9-10980XE. 

Intel often has to lean on its overclocking advantage to wrest the lead from the 3950X, but the 3950X isn't too shabby after overclocking, either. Intel's advantage after tuning comes at a $230 premium and requires more robust cooling and power delivery, so you should factor that into your purchasing decision. 

We generally don't recommend HEDT parts for gamers, you'll get faster performance from the mainstream Core i9-9900K and 3950X, but streamers do benefit from the extra cores and threads, particularly for editing tasks. Intel's new $979 price point is much more palatable for this class of user, but most of the benefit over the 3950X would only come after spending those extra dollars on accompanying components. Most would be better served by the Ryzen 9 3950X. 

If you don't need extra I/O or quad-channel memory, the 3950X is also a better value for most productivity workloads. That does leave a preciously slim slice of the market where Intel has an advantage in this price bracket (users that need quad-channel memory or more PCIe lanes). Overclocking performance is a factor if you're willing to spend the cash. You can drop the -10980XE into an existing X299 board if you're willing to sacrifice a few PCIe lanes, but be aware that this is the end of the line for the X299 platform. 

Meanwhile, AMD's Threadripper 3000 platform will likely be compatible with future generations, and we suspect the company will soon release a Threadripper part to compete directly with the -10980XE.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Software vendors got some bad news over the weekend. ZDNet reported that Chinese hackers demonstrated exploits in major browsers, common utilities and other apps during the Tianfu Cup hacking competition held in Chengdu.

China has officially "discouraged" security researchers from participating in hacking competitions outside the country since at least March 2018. Tianfu Cup gives those hackers a place to demonstrate their skills--and earn six-figure bounties for successful exploits--without risking the government's ire by competing abroad.

ZDNet reported that former Pwn2Own winner Team 360Vulcan earned $382,500 for hacking the old version of Microsoft Edge, Office 365, Adobe PDF Reader, qemu+Ubuntu and VMWare Workstation across the two days of competition. Some $200,000 of that came from the VMWare exploit; another $80,000 came from the qemu+Ubuntu exploit. The remaining $102,500 was split among the other apps.

Team 360Vulcan wasn't successful in all its attempts. An exploit targeting iOS that was supposed to close out the competition reportedly didn't go as planned. Both days of competition were a bit of a mixed bag, actually, with roughly half of the exploits working as intended (that's 13 of the planned 32 for the first day and seven of the planned 16 for the second day). Maybe there's solace to be found there.

Teams also demonstrated successful exploits in Chrome, Safari and the D-Link DIR-878 router during the Tianfu Cup's two-day competition. Earnings varied based on the severity of the vulnerability and the company that made the product. And don't worry, because the competition's organizers told ZDNet that all teams plan to reveal details about the exploited vulnerabilities to the affected companies after the event.

It would be easy to demonize events like this. Some of these products are used by hundreds of millions of people; undermining their security for sport might seem a bit strange. But these competitions help incentivize ongoing research into these products that companies then use to make them even more secure. It's better to have public displays of security exploits than to for them to be used secretly with malicious intent.

SSDs are dirt cheap because the market is flooded with excess flash, new technologies QLC (quad-level cell) have made memory cheaper than ever and companies are getting more aggressive with pricing. Prices have declined so drastically that you can now supercharge your PC with 500 GB of flash for well under $70 or under $100 for 1TB.  

Made for price-conscious buyers who need 2.5-inch SATA drives, Crucial BX500 performs competently, but doesn't even come close to being one of the best SSDs. With so many superior products in the same price range, it's hard to recommend.

Crucial's BX500 is the successor to its popular BX300 line of SSDs. Like the company's mainstream MX500 brand, the BX500 skips over the 400-naming scheme. But, unlike the MX500, the BX500 doesn't offer much of an upgrade path over its predecessor.

The BX series is a streamlined, no-frills SSD with fewer accessories and features than the MX series. Crucial launched the BX series to tempt buyers into purchasing flash when other options in the market were too expensive. The original BX100 came with 16nm planar (2D) MLC flash and a Silicon Motion (SMI) controller. That was Crucial’s first SSD with an SMI controller, and that trend continues with the BX500. The SSD uses the new SM2258XT four-channel SSD controller paired with Micron’s latest 64-Layer 3D TLC flash.

The base SM2258 is a good SATA SSD controller, and it offers impressive performance and reliability if it’s paired with the right flash. But it needs an expensive DRAM package for caching.

The SM2258XT, known as the XT model, combats this by removing the need for DRAM. This allows the SSD to store the critical flash translation layer directly on the flash instead of in a DRAM buffer. This lowers prices by a few dollars, but it also results in lower performance. NAND isn't as fast as DRAM, and the constant read/write modifications to the flash translation layer are a strenuous task. As a result, performance can be rather unflattering–even falling into hard drive territory.

Controller and 96-Layer Flash

If it goes on a huge sale, Crucial’s BX500 might be a decent choice if you’re in search of a low-cost SSD to store your games library. While it’s not a top performer, it outpaces HDDs with speeds of up to 540/500 MB/s of read/write throughput. It's also pretty cheap with a price of just $0.10 cents per GB, though it isn't quite as cheap as some of its competition.

Flash pricing has plummeted drastically in the past year, making higher-capacity SSDs much more affordable. You can now buy 1TB SSDs for as little as $100 (or less if you happen to snag a sale). This has driven up demand for higher capacities so much that Crucial released a 960GB model for its BX500 line.

The 960GB BX500 is not the fastest drive on the market, but Micron’s latest 96-Layer 3D TLC flash and a new SM2259XT controller help make it affordable. There are also 1TB and 2TB capacities that we didn't get to test.

Specifications

Crucial’s BX500 provides up to 540/500 MB/s of sequential read/write throughput, but that can drop to an average of just 100 MB/s during a sustained workload. Crucial doesn't disclose random 4K IOPS performance, likely due to unimpressive performance, but we'll measure it on the following pages.

The BX500 is available in 120GB, 240GB, 480GB, 960GB, 1TB and 2TB capacities. The BX500 currently sells for ~$0.10-per-GB for the 960GB and 2TB models.

A three-year warranty and affordable prices designate the BX500 an entry-level SSD, but its rather low write endurance epitomizes its rank. The BX500's endurance rating starts at 40 TBW (Terabytes Written) and spans up to 120 TBW. Those are among the lowest endurance ratings on the market. Surprisingly, the BX500's endurance is even lower than its predecessor.

Crucial’s BX500 comes in a 2.5” 7mm form factor and communicates with the host system via a SATA 6Gb/s link. The 960GB model comes with the same performance rating of up to 540/500MB/s read/write. The drive features a 3-year warranty but has twice the endurance (up to 240TBW).

Accessories

Crucial includes Acronis True Image HD and Crucial’s own SSD toolbox, Crucial Storage Executive, with the BX500 SSDs.

True image enables end users to quickly and safely migrate their data from their old drive to their new BX500. You can also perform system backups with it, too. Crucial’s Storage Executive is also a handy tool. With it, you can update your firmware, monitor your SSD, and enable momentum cache, which Crucial states can help improve performance up to 10x in some cases.

A Closer Look

The BX500’s case consists of thin metal and plastic. That keeps it lightweight, but the plastic gives it a cheap feel, which stands in stark contrast to its predecessor and the MX series. It connects to the host via a SATA 6GB/s connection.

Taking the casing apart reveals a 1/4 sized PCB, which is another way to reduce costs while still keeping compatibility with the 2.5" form factor. The SM2258XT resides near the connector for the best signal, and the 64-Layer 3D TLC flash is distributed among four emplacements (two on each side). Raw NAND capacity is 51GB, but after over-provisioning, the user addressable space is just 480GB. That drops to 446GB of addressable space after you format the drive in Windows.

The BX500 features a plastic casing that simply snaps together to hold the small 1/4 sized PCB within. This helps to reduce weight and material cost but leaves it with a low-quality feel. And, unlike most other SSDs, there are no thermal pads to help dissipate heat from the controller.

Four Micron 96-Layer TLC flash packages are distributed in pairs on each side of the PCB. After factory provisioning, the 960GB model leaves you with a total usable space of 894GB within Windows.

The four-channel SMI SM2259XT controller sits near the SATA connector. While similar to the previous SM2258XT, this model comes with some data path improvements to boost performance.

The Silicon Motion SM2259XT SATA controller uses a DRAM-less architecture, so there is no need for a DRAM package. Instead, the critical mapping information is stored and modified on the flash. While this helps reduce BOM cost and allows for lower prices overall, it can drastically reduce performance, which you will see on the following pages.

240GB Comparison Products

We included the Intel SSD 545s, Samsung 860 EVO, Crucial MX500, Plextor M8V, Toshiba OCZ TR200, and WD Blue 3D as comparisons in this review. All feature 3D TLC flash and current-gen controllers optimized for consumer workloads. We have also added results from a 960GB Intel Optane SSD 905P and a 2TB WD Blue HDD in certain tests.

Trace Testing – PCMark 8 Storage Test 2.0

PCMark 8 is a trace-based benchmark that uses Microsoft Office, Adobe Creative Suite, World of Warcraft, and Battlefield 3 to measure the performance of storage devices in real-world scenarios.

The BX500 has the lowest performance of our comparison pool. As we can see, the DRAMless architecture is a definite weakness in tests that replicate real-world applications that tend to have large datasets.

Game Scene Loading - Final Fantasy XIV

The Final Fantasy XIV StormBlood benchmark is a free real-world game benchmark that easily and accurately compares game load times without the inaccuracy of using a stopwatch.

In this test, most of the top-performing SSDs achieve a total load time of about 22 seconds. The BX500's 28-second load time is better than the HDD, but still much slower than the fastest SSDs. 

Transfer Rates – DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with our own custom 50GB block of data. Our data set includes 31,227 files of various types, like pictures, PDFs, and videos. We copy the files to a new folder and then follow up with a read test of a newly-written 6GB file.

Crucial’s BX500 is unimpressive during our copy test. Again, it falls into last place against the other SSDs. The BX500's average speed of just 68 MB/s is a slim 21 MB/s faster than the HDD. The BX500 averaged 498 MB/s during the read test, which is slightly faster than the WD Blue 3D SSD and a huge advantage over the HDD.

SYSmark 2014 SE

Like PCMark, SYSmark uses real applications to measure system performance. SYSmark takes things much further, however. It utilizes fourteen different applications to run real workloads with real data sets to measure how overall system performance impacts the user experience. BAPCo's SYSmark 2014 SE installs a full suite of applications for its tests, which includes Microsoft Office, Google Chrome, Corel WinZip, several Adobe software applications, and GIMP. That also makes it a great test to measure the amount of time it takes to install widely-used programs after you install a fresh operating system.

SYSmark installed on the BX500 three and a half minutes faster than it did on the HDD, but it still served up low performance during the application performance tests. The BX500 provided more than twice the performance of the HDD, but it was still the slowest SSD in the test pool.

ATTO

ATTO is a simple and free application that SSD vendors commonly use to assign sequential performance specifications to their products. It also gives us insight into how the device handles different file sizes.

The 240GB BX500 delivers the typical throughput performance we expect out of a SATA SSD. In this test, it matched its 540/500 MB/s of read/write throughput specification.

Anvil's Storage Utilities

Anvil's Storage Utility is a commonly-referenced benchmark that simplifies the complex IOMETER benchmark and its underlying Dynamo engine with a one-click software wrapper.

Crucial’s 240GB BX500 achieved a low result during the read test, but it beat the M8V and WD Blue 3D during the write test. The mismatched read and write performance leads to a total score that ranks second to last.

CrystalDiskMark

CrystalDiskMark (CDM) is a simple and easy to use file size benchmarking tool.

The BX500 reached 560/494 MB/s of peak sequential read/write throughput at QD (queue depth) 32. At a QD of 1, it achieves the second-fastest result during the read test, but again, the lowest result during the write test. The drives' 4K random performance, which is a key measure of snappiness in real-world applications, again proves to be lower than the norm.

Sustained Sequential Write Performance

Official write specifications are only part of the performance picture. Most SSD makers implement an SLC cache buffer, which is a fast area of flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the SLC cache and into the "native" TLC or QLC flash. We hammer the SSDs with sequential writes for 15 minutes to measure both the size of the SLC buffer and performance after the buffer is saturated.

Just like many other SSDs, the BX500 utilizes an SLC cache to absorb small bursts of write traffic. After about 13GB of continuous writes, performance degrades from an average of 500 MB/s to 100 MB/s.

Power Consumption

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is a very important aspect to consider, especially if you're looking for a new drive for your laptop. Some SSDs can consume watts of power at idle while better-suited one's sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state faster, which ultimately saves power.

The 240GB BX500 manages its power consumption rather well. It consumes just 27mW with LPM enabled, and 0.4W with the feature disabled. Both results land in third place.

The 240GB BX500 continues to sip power during the file copy test. It averages 1.55W, which is the second lowest result. Crucial’s BX500 may sip power at idle and under load, but that doesn’t mean it is efficient. The BX500's average of 44 MB/s-per-watt lands in last place, which is surprising given the DRAMless architecture.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

480GB Comparison Products

Trace Testing – PCMark 8 Storage Test 2.0

PCMark 8 is a trace-based benchmark that uses Microsoft Office, Adobe Creative Suite, World of Warcraft, and Battlefield 3 to measure the performance of storage devices in real-world scenarios.

The 480GB BX500 is even slower in this test than the 240GB model–it lags significantly behind the other drives and lands in last place.

Game Scene Loading - Final Fantasy XIV

Most 500GB-class SATA SSDs load the Final Fantasy game level in 22-23 seconds. The 480GB BX500 isn't too much slower with a 24.5 second load time, but it still trails the other SATA SSDs. It does, however, offer significantly improved game load performance over the HDD, making it a viable option for storing your games library. 

Transfer Rates – DiskBench

Crucial’s BX500 is unimpressive in our real-world copy test. Again, it falls into last place against the other SSDs. Its average of just 58 MB/s is just an 11 MB/s faster than an HDD.

The BX500 read data at an average speed of 496 MB/s, landing just ahead of the WD Blue 3D SSD.

SYSmark 2014 SE

SYSmark installed onto the BX500 faster than our HDD by about three and a half minutes, but that was still a minute slower than any other SSD in our comparison pool. The BX500’s poor performance continued during the application testing.

ATTO

Like the 240GB model, the 480GB BX500 surpassed its sequential read specifications.  

Anvil's Storage Utilities

Crucial’s BX500 ranks behind our comparison pool once again. It displays strong sequential performance, but performance in random workloads holds it back.

CrystalDiskMark

The BX500 reached sequential read/write speeds of 557/492 MB/s at QD32. These aren't world-beating numbers, but they are more than acceptable for a SATA SSD. The drive scored the fastest QD1 sequential read result in our test pool, but the lowest result in the sequential write test.

Random performance isn't very promising. The BX500 ranks in last place again at QD32. The drive reads data at just 31 MB/s at QD1, making it the slowest in the group, but its 107 MB/s of random write performance ranks second. 

Sustained Sequential Write Performance

The larger 480GB model can write more data continuously than the 240GB model before it slows down. Performance degrades to just 100 MB/s after about 24GB of continuous writes.

Power Consumption

The BX500’s sips just 31mW at idle with LPM enabled, and 0.4W when we disable the feature. Both results rank in second place, just behind the Samsung 860 EVO.

The BX500 averages 1.42W during the file copy test, which is the best in our comparison pool. The BX500's peak consumption comes in second.

Like the 240GB model, the 480GB BX500 isn’t that efficient. It averaged 44 MB/s-per-watt during the file copy, again trailing the other SSDs.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

960GB Comparison Products

We have added a few mainstream options to our comparisons for our 960GB BX500 testing. These include Crucial’s own MX500, Samsung’s 860 QVO and EVO, as well as WD’s Blue 3D and Intel’s SSD 660p. We have also added results from a 960GB Intel Optane SSD 905P and a 2TB WD Blue HDD for reference in a few of the real-world workloads.

Trace Testing – PCMark 8 Storage Test 2.0

PCMark 8 is a trace-based benchmark that uses Microsoft Office, Adobe Creative Suite, World of Warcraft, and Battlefield 3 to measure the performance of storage devices in real-world scenarios.

Crucial’s 960GB BX500 scores a total of 4580 points and an average bandwidth of 157 MB/s. This ranks as one of the lowest-performing SSDs we've tested and lands in last place. While it will be suitable for light workloads, it can’t keep up with the rest of the mainstream options under slightly tougher conditions.

Game Scene Loading - Final Fantasy XIV

The Final Fantasy XIV StormBlood benchmark is a free real-world game benchmark that easily and accurately compares game load times without the inaccuracy of using a stopwatch.

Loading Final Fantasy’s game scenes proved to be a task for the BX500. While it features similar read performance as the other drives, the BX500 was 2-5 seconds slower. Performance is still much better than any HDD, however.

Transfer Rates – DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with our own custom 50GB block of data. Our data set includes 31,227 files of various types, like pictures, PDFs, and videos. We copy the files to a new folder and then follow up with a read test of a newly-written 6 GB file.

Crucial’s 960GB BX500 provided great read performance during our real-world copy test. It read our 6GB test file at an average rate of 494MB/s, which is in line with most other SATA SSDs.

The BX500 fell on its face during our 50GB file transfer. With an average of 49 MB/s, the BX500’s copy performance proves to be rather pitiful, at best.

SYSmark 2014 SE

Like PCMark, SYSmark uses real applications to measure system performance. SYSmark takes things much further, however. It utilizes fourteen different applications to run real workloads with real data sets to measure how overall system performance impacts the user experience. BAPCo's SYSmark 2014 SE installs a full suite of applications for its tests, which includes Microsoft Office, Google Chrome, Corel WinZip, several Adobe software applications, and GIMP. That also makes it a great test to measure the amount of time it takes to install widely-used programs after you install a fresh operating system.

Just as we have seen during previous tests, the larger BX500’s performance isn’t much different during SYSmark. The 960GB BX500 installed SYSmark faster than an HDD by about 30 seconds, but other SSDs accomplished the task 3-4 minutes faster overall. The drive notched a responsiveness score of 1552 points, which nearly matches the Samsung 860 EVO and surprisingly surpasses the Samsung 860 QVO.

ATTO

ATTO is a simple and free application that SSD vendors commonly use to assign sequential performance specifications to their products. It also gives us insight into how the device handles different file sizes.

The BX500's performance in ATTO is similar to other SATA SSDs. Here the BX500 matched its ratings of 540/500 MB/s of read/write throughput. However, we see there is a dip in write performance that spans the 128KB-512KB file sizes. This is due to the DRAM-less architecture, which causes the BX500 to “get into a condition or state where background garbage collection briefly has to take precedence over I/Os” during the test.

Anvil's Storage Utilities

Anvil's Storage Utility is a commonly-referenced benchmark that simplifies the complex IOMETER benchmark and its underlying Dynamo engine with a one-click software wrapper.

Crucial’s 960GB BX500 delivered low scores across the board in the Anvil benchmark. It achieved a total score of 4,282 points, which ranks in last place.

CrystalDiskMark

CrystalDiskMark (CDM) is a simple and easy to use file size benchmarking tool.

Crucial’s 960GB BX500 reached 562/337 MB/s of write performance at a queue depth (QD) of 32. At a QD of 1, it reached 504/442 MB/s of read/write throughput. Looking at random performance, the BX500 scored the lowest read and write scores. During heavier random access it hit just over 50,000 read IOPS, and a pitiful 1,875 write IOPS.

This goes to show that the BX500, again, hit a situation where background activity took precedence over delivering the IOs the host requested.

Sustained Sequential Write Performance

Official write specifications are only part of the performance picture. Most SSD makers implement an SLC cache buffer, which is a fast area of SLC-programmed flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the SLC cache and into the "native" TLC or QLC flash. We hammer the SSDs with sequential writes for 15 minutes to measure both the size of the SLC buffer and performance after the buffer is saturated.

As expected, the 960GB BX500 uses an SLC cache for burst write performance of up to its 500 MB/s rating. After 48GB of data written, performance degrades to a write speed of 85MB/s. While this allows it to keep up with our comparison pool for the initial burst, most other SSDs will easily surpass its performance during long sustained write workloads.

Power Consumption

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is a very important aspect to consider, especially if you're looking for a new drive for your laptop. Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state faster, which ultimately saves power.

At idle the BX500 consumes fairly low power. Without LPM enabled, it consumes just under half a watt. With it enabled, it sips just 0.028W.

The drive consumes the least amount of power in our comparison pool under load, and the maximum power draw is similar to the Samsung 860 EVO. But, while it consumes the least amount of power, the slow transfer speed of 49 MB/s during the 50GB copy test leads to an average of just 39 MB/s per watt. The DRAM-less architecture lands it in the last place for efficiency.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

After Micron’s 96-Layer TLC flash became available, Crucial took the opportunity to add in a 960GB model to the BX500 SSD series and then the 1TB and 2TB (which we haven't tested). Crucial also moved the higher capacity drives to a newer, SM2259XT controller.

While Crucial's transition to the new 96-Layer flash and improved controller were welcome, it didn't  drastically change performance for the better. Looking through both the 480GB and 960GB results, we see a slight decrease in performance in some situations, like sustained write performance. The BX500 averaged 85 MB/s of sustained write speed after the SLC cache filled, so we see the decreased performance in our 50GB copy transfer test. Power efficiency suffers as a result.

The majority, if not all, of the similarly-priced (or cheaper) drives outperformed the 960GB BX500. The BX500 did well in SYSmark, but when we factor in the overall performance results, the result is a rather underwhelming SSD.

The BX500 does have a low price point, though. At $99, or $0.10 per GB, it is very affordable. However, it is only $7 cheaper than its much faster brother, the MX500. The MX500 also comes with a higher endurance rating. It’s well worth the extra few dollars for the increased performance and endurance.  Considering the performance delta, it's not worth saving that $7.

Meanwhile, the Intel SSD 660p is a few dollars cheaper and offers much more performance across the board. That is due to it being an M.2 NVMe SSD, so it is not a direct comparison if you can’t use an M.2 SSD in your system.

Be sure to weigh your options according to what you value most. Your hard-earned dollars could be better spent or saved by looking to alternatives.

Crucial 240GB and 480GB Conclusion

Crucial’s BX500 has a cost-reducing DRAMless architecture, but that has a big impact on performance. Even without DRAM, the SSD can achieve the typical sequential throughput we expect from a SATA SSD, but the BX500's all-important 4K random performance, a key measurement of drive snappiness, is lower than most SSDs.

We have seen this type of performance before from drives like Toshiba's TR200, which also doesn’t have a DRAM cache. Like the BX500, that drive exhibits low performance in random workloads. Write performance also suffers during extended write workloads.

As we saw in the PCMark 8 and SYSmark 2014 benchmarks, the Crucial BX500 is unimpressive in tests that measure real-world application performance. The BX500 surpasses an HDD, without question, but there is a clear divide between it and other SSDs. While it comes with an SSD toolbox and Acronis True Image HD for migrating your data, the BX500's value proposition still leaves a lot to be desired.

The BX500 family is priced competitively at $0.15- to $0.22-per-GB, but there are better options for only a few dollars more. It appears the viability of lower-tier products has nearly vanished at these low capacities. The BX500 would need to have a significantly smaller price tag for us to recommend it. Even if you are pinching pennies, we would recommend the WD Blue 3D, MX500, or Samsung 860 EVO. All those drives boast longer warranties, greater performance, and endurance at competitive price points.

Image Credits: Tom's Hardware

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

Adobe Flash will officially keel over in 2020. Its presence on the web might be forgotten even sooner than that, however, because Google announced yesterday that its search engine will stop indexing Flash content by the end of the year.

The question used to be "if a tree falls in the forest but nobody's around to hear it, does it make a sound?" Now, it's "if something exists on the Internet but Google doesn't index it, does it really exist?" Search has become the default way of finding stuff online--Flash's disappearance from Google's results basically makes it invisible.

Here's the good news: Google isn't removing websites that have Flash content from its search results. That would be a disaster, because people who operate sites that still rely on Flash probably wouldn't be technically savvy enough to understand why they've disappeared from Google. The pages containing Flash content will remain.

Instead, "in Web pages that contain Flash content, Google Search will ignore the Flash content," Google explained. Google will also "stop indexing standalone SWF files." But the search tool assured us that "most users and websites won't see any impact from this change" to Google Search.

This isn't Google's only method of helping Flash along its journey into the afterlife. The plugin's disabled by default in Chrome, and the company plans to completely remove support for Flash from its browser by the end of 2020. (These changes also affect the Chromium-based version of Edge that Microsoft's been developing.)

Many have acknowledged that Adobe ending Flash support marks the end of an era. Google said in its blog post that the plugin "was the answer to the boring static web, with rich animations, media and actions" and was "a prolific technology that inspired many new content creators on the web." But now it's time for Flash to fade away.

Adobe said earlier this month that U.S. sanctions would force it to suspend its Creative Cloud service in Venezuela. Its customers in Venezuela got some good news today, however, because the company announced it's received permission from the U.S. government to continue offering its service in the country, despite the restrictions.

Adobe originally said it had to deactivate accounts in Venezuela because of Executive Order 13884. The vendor told users in Venezuela that they'd have until October 28 to download files managed via Creative Cloud, Lightroom, Document Cloud or Adobe Spark; otherwise the documents would be deleted. 

A week later, Adobe said that it would be able to offer refunds, provided the subscription was purchased directly from its website. It also told us that it was "working with our partners on the same" arrangement. Adobe would continue to operate Behance, a social platform for creative professionals, in Venezuela as well. 

Now it's published a blog post about its ability to stay in Venezuela.  

"After discussions with the U.S. government, we’ve been granted a license to provide all of our Digital Media products and services in Venezuela. With this update, we’re sharing that users can continue to access the Creative Cloud and Document Cloud portfolio, and all of their content, as they did before. If you lost access to premium services, they will be restored within a week," it says.

Adobe also updated its FAQ article regarding accounts in Venezuela. The company said subscribers who received refunds will be given free access to the same services for 90 days and that it "will contact you mid-November with any steps you need to take to renew your subscription." There shouldn't be any interruption to its services, and Adobe didn't put an end date on its agreement with the U.S. government.

Updated, 10/14/19, 9:15am PT: Adobe sent us the following statement:
"We can confirm that Behance will continue to be accessible in Venezuela. And, Customers who purchased directly from Adobe will be refunded by the end of the month. We are working with our partners on the same. We regret the difficulties this causes our customers. We will share more details about how our operations and customer activities might be impacted, as they become available."

Behance is the company's social platform for creative professionals. We asked Adobe what prompted this change in its response to Executive Order 13884. It's not clear if the company's understanding of the order changed, if it's capitulating to backlash or if it was given a special exemption to issue refunds in Venezuela despite the restrictions.  The company has not responded to our request for additional information.

Original article, 10/8/19, 8:19am PT:

Change is hard. Adobe's switch to subscription-based pricing was no exception, forcing users to pay a monthly fee to use Photoshop, Illustrator and the software vendor's other creative tools, rather than being able to buy them outright. Those people are likely feeling some amount of vindication since Adobe yesterday revealed that it has to deactivate the accounts of Venezuelan customers to comply with U.S. sanctions.

The U.S. government issued Executive Order 13884 on August 7 to prohibit U.S. companies from doing business in Venezuela. That means Adobe has to deactivate these accounts, and even though it promised to "continue to monitor developments closely" and "make every effort to restore services to Venezuela as soon as it is legally permissible to do so," the company has no way of knowing when the sanctions will be lifted.

But that's not even the worst part. Adobe said in the support document that it's "unable to issue refunds" because Executive Order 13884 "orders the cessation of all activity with the entities including no sales, service, support, refunds, credits, etc." People who signed up for Adobe's services with an annual subscription, then, effectively gave the company a bunch of money for services it can no longer offer because of U.S.-Venezuela relations.

Adobe customers, (which includes users of services like Creative Cloud, Lightroom, Document Cloud and Adobe Spark), in Venezuela have until October 28 to download any files. Afterwards, their Adobe accounts will be deactivated. Adobe didn't clarify whether that means the stored files will be permanently deleted on that date or if it will hold onto them in the hopes of the U.S. lifting sanctions on Venezuela.

It can be easy to justify software companies switching to subscription offerings (just think of the Microsoft Office - Office 365 debate). Most regularly update their services, and the subscription model encourages them to keep working on those updates rather than focusing on major upgrades that have to be purchased separately. A lot of services also include some amount of cloud-based storage that incurs extra costs for the company that are then passed on to the consumer via the monthly fee.

But the looming deactivation of these accounts represents an irrefutable argument against subscription offerings: people are paying for something that can be taken away from them at any time. Venezuelans who have an older version of Adobe's products won't see them forcibly removed from their PCs. Venezuelans who embraced the company's new business model, however, might as well have burned their money instead.

Researchers from Ruhr University Bochum and Müns­ter Uni­ver­si­ty revealed PDFex, two vulnerabilities of PDF files that undermine the encryption used to secure their contents. One vulnerability lets attackers manipulate parts of the file to enable direct exfiltration attacks, and the other can be used to "modify existing plaintext" and "construct entirely new encrypted objects."

The first vulnerability works because "the PDF specification allows the mixing of ciphertexts with plaintexts," the researchers explained, and potential attackers could then use "further PDF features which allow the loading of external resources via HTTP" to steal the file's contents. They managed to exfiltrate data from an encrypted PDF via PDF forms, hyperlinks and JavaScript code added to the original document.

The second vulnerability results from PDF encryption's use of "the Cipher Block Chaining (CBC) encryption mode with no integrity checks, which implies ciphertext malleability." This lets attackers "create self-exfiltrating ciphertext parts using CBC malleability gadgets." Attackers can use the same methods of exfiltration--PDF forms, hyperlinks and JavaScript--to access file contents after exploiting this vulnerability.

This isn't an isolated problem. The researchers explained that many companies rely on PDF encryption. Some, like Canon and Samsung, use PDF encryption in their scanners. IBM offers "PDF encryption services for PDF documents and other data (e.g., confidential images) by wrapping them into PDF," they said, and PDF encryption can also be used to keep medical records secure during transfer. 

The PDFex vulnerabilities are also hard to avoid because they're problems with the PDF format itself. The researchers said their "evaluation shows that among 27 widely-used PDF viewers, all of them are vulnerable to at least one of those attacks, including popular software such as Adobe Acrobat, Foxit Reader, Evince, Okular, Chrome, and Firefox." They shared more information about this evaluation on a dedicated website.

Lavish and Spacious

LaCie's simply named Mobile Drive is a sleek-looking portable HDD, for those looking for some style in their external storage lives. But that’s really all it has going for it compared to most other comparable external drives. In testing, it hit speeds of up to 140MBps, which isn’t too bad for a 2.5-inch HDD-based external, but is in the stone age compared to SSD-based alternatives. And, although it does come in at lower cost than flash, be prepared to pay the LaCie tax for its aesthetics. LaCie demands a significant premium, at $25-50 per capacity, over similarly performing external HDD options on the market.

LaCie’s latest generation of external HDDs focuses on lavish design. And we can't blame the company, given that single-drive portable disk-based storage tends to be pretty similar (and stagnant) when it comes to performance, making it tough to differentiate. This portable external HDD feels as solid as it looks. The diamond-cut aluminum finish makes this drive one of the classiest-looking devices we have used this year. But, looks aren’t all it has going on.

LaCie’s Mobile drive comes in spacious capacities to store all your data. LaCie states that their largest capacity can store up to 165 hours of 4K video or up to 500,000 photos. It utilizes the new USB-C interface, but, because it’s still an HDD inside, can’t begin to saturate the interface.

Specifications

LaCie lists the availability of the Mobile Drive in four capacities: 1TB, 2TB, 4TB, and 5TB. All don’t have an official rating on LaCie’s website, but in testing, they can hit speeds of up to  about 140 MB/s. Unlike their SSD counterparts, the Mobile Drive features a shorter 2-year warranty.

Software and Accessories

Included with the drive are two USB cables, one a USB Type-C to Type-C cable and the other a USB Type-C to USB Type-A cable for broad compatibility. This means the drive will work with Thunderbolt 3- and USB 3-compatible devices.

From the factory, LaCie’s Mobile Drive comes pre-formatted as exFAT so that it is both PC and Mac compatible. Pre-loaded on the drive are a warranty document and a quick-start application to help you set up your Adobe membership and download LaCie’s Toolkit, which can be used to back up your data.

Like the LaCie Rugged RAID Pro we reviewed a few months ago, LaCie's Mobile Drive also features a 1-month membership to Adobe's Creative Cloud All Apps plan ($53 if purchased separately). Unfortunately, there is no encryption feature or software included.

Closer Look

LaCie's Mobile Drive features an all-aluminum sandblasted finish and diamond-cut design in moon silver color. In addition to moon silver, there is also a space gray model to fully match your Mac, but it is only sold by Apple at this time.

Our 2TB review sample measures in at 10 x 87.8 x 121.6 mm and weighs about 74g. The 4TB and 5TB models are double the thickness and slightly larger overall. These also weigh about double what the 1TB and 2TB models weigh, due to additional storage platters on the internal drives.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: Best External Hard Drives and SSDs

Performance Results

Comparison Products

To gauge the performance of the LaCie Mobile Drive, we’ve opted to pit it against a few other external SSDs and HDDs that use Thunderbolt 3 and various USB interfaces. Included are results for the 4TB Rugged RAID Pro, which utilizes two 2.5” HDDs in RAID 0. Next, we’ve taken out the 1TB SanDisk Extreme Portable SSD and 480GB Kingston HyperX Savage EXO, both SATA, SSD-based competitors. As well,, we included some Thunderbolt 3-based competition to stir things up a bit, just to show how fast these SSDs really are over HDD competition. These included a 1TB G-Technology G-Drive mobile Pro and 1TB Samsung X5.

Transfer Rates – DiskBench

DiskBench is a storage benchmarking tool that allows us to test the transfer or copy performance of a storage device with real data. We test external drives with three file transfers that consist of 25GB of photos (10GB of iPhone jpgs and 15GB of RAW photos from a Canon 6D), 50GB of movies, and 25GB of documents. First, we transfer each folder from a 1TB NVMe SSD to the external device; then we follow up by reading a 3.7GB 7-zip file and a 15GB movie back from the device.

When it comes to moving real files, LaCie’s Mobile Drive did pretty well considering its slow 2.5-inch HDD within. It came in the last place across the board in both writing in reading. It wrote our photos folder at 105 MBps, the movies folder at 132 MBps, and our docs folder at 91 MBps. Reading back large files resulted in an average speed of about 120 MBps.

Trace Testing – PCMark 8 Storage Test 2.0

PCMark 8 is a trace-based benchmark that uses Microsoft Office, Adobe Creative Suite, World of Warcraft, and Battlefield 3 to measure the performance of storage devices in real-world scenarios.

In PCMark 8, LaCie’s Mobile drive scored 1,585 points in total and averaged just 5 MBps, plopping it into last place. If you are for a drive in your professional workflow for anything other than backup, we suggest looking elsewhere to flash-based SSD alternatives.

Synthetics - ATTO

ATTO is a simple and free application that SSD vendors commonly use to assign sequential performance specifications to their products. It also gives us insight into how the device handles different file sizes.

Just as in PCMark 8, the LaCie Mobile Drive ranks last on our ATTO charts. Sequential speeds hit nearly 140 MBps in both read and write.

Synthetics - iometer

iometer is an advanced and highly configurable storage benchmarking tool that vendors often use to measure the performance of their devices

LaCie’s Mobile Drive falls into the last place on our charts once gain. It maxes out at  roughly 140 MBps read/write. Random read latency comes in at 8ms at best, while write latency is much lower at just 3ms. While these numbers are decent for a hard drive, they pale in comparison to something like the HyperX Savage EXO and SanDisk Extreme SSDs.

Write Saturation and Temperature

Write speed and temperature are two important and inter-related metrics for external devices. We threw in this one last test to measure the performance of the drive over a 15-minute window. When possible, we also log the temperature of the drive via the S.M.A.R.T. data to see when (or if) thermal throttling kicks in and how it impacts performance. Bear in mind that results will vary based on the workload and ambient air temperature.

Sequentially writing to the drive results in the slowest rate out of the test group, taking 15 minutes to write just 84GB. Over the course of writing the data, we logged the highest temperature of 38C, which isn’t too hot at all.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

Conclusion

LaCie’s Mobile Drive is more aesthetically pleasing than your average portable HDD. Its diamond-cut edging and sandblasted aluminum finish rank it up there as one of the best-looking and -feeling external HDDs we have come across to date. It's definitely a top pick if you want to match your external drive to your Mac or other aluminum-finished devices.

But flash-based storage is still light years ahead in terms of speed, and prices of SSDs have been steadily and dramatically falling over the last few years. So if you care about performance, you should probably opt for an external SSD. Performance of the Mobile Drive is average for an HDD of this caliber. In testing, we hit speeds of up to 140 MBps read and write and averaged transferring data around at about 120MBps. If you plan on doing anything other than backing up, your experience may be a bit laggy. Flash-based portable drives are much better for any more-demanding tasks, as demonstrated by our PCMark 8 results.

We appreciate that LaCie included USB-C to USB-A and USB-C to USB-C cables for broad market compatibility. Not everyone has the latest and greatest devices with USB-C or Thunderbolt 3 support. And, LaCie’s Toolkit is a great software tool for end-users to utilize for their backup routine. It's quite simple to use and gets the job done without much effort. It takes the guesswork out of backing up.

One thing that we think LaCie could improve is adding a rubber texture to the bottom of the drive to prevent slipping on desks and other surfaces. Other than that, the drive is a pretty solid build. The only other qualm we have is the premium price being so much higher than competing portable hard drives, to the point that you can get double the storage capacity, plus similar performance and warranty for roughly the same price or just a little more. If you are in search of a new portable HDD, LaCie’s Mobile Drive wouldn’t be our first recommendation, especially if you care about value. But if looks matter to you, externals don’t get much prettier than LaCie’s Mobile Drive.

Image Credits: Tom's Hardware

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: Best External Hard Drives and SSDs

Memory Overclocking on X399 Colfax

A fine line has always separated the HEDT (high-end desktop) platforms from the mainstream platforms. You commonly find hardcore enthusiasts, content creators, and workstation users opting for the first, while the average computer user or gamer often settle for the latter. Since the HEDT platform targets more demanding users, it naturally comes with a bigger feature set, such as support for core-heavy processors or a more generous amount of memory slots, storage options, and PCI Express slots.

Sadly, you don't get the extra features for free, though. Building a PC around a HEDT platform typically requires a bigger investment because HEDT parts are considerably more expensive. You probably would want to spend your hard-earned money where it matters most. This article will evaluate the impact that memory speeds have on the X399 platform. Our testing includes multiple memory speeds that span from DDR4-2133 up to DDR4-3333 in different scenarios to determine whether the benefits of investing in the faster memory kits outweigh the cost.

Colfax and DDR4-2933

The AMD X399 chipset, which is a little over two years old now, continues to be the Red Team’s flagship enthusiast platform. Since its inception, the X399 chipset has seen and housed up to two generations of AMD Ryzen Threadripper desktop processors. As one would expect, memory support has gone through its own gradual evolution.

Memory support on AMD's Ryzen Threadripper processors has progressed like the Ryzen desktop chips. The first-generation Threadripper chips, codenamed Whitehaven, debuted with support for DDR4 memory modules up to 2,666 MHz. Memory support wasn't bad, but it wasn't spectacular either. On the other hand, the second-generation Threadripper parts, codenamed Colfax, arrived with support for memory speeds that escalate to 2,933 MHz, which represents a 12.5% improvement over the first-gen Threadripper processors.

One of the selling points for the Ryzen Threadripper processors is the native support for quad-channel memory. As a result, AMD X399-based motherboards that adhere to the ATX form factor frequently come equipped with eight DDR4 memory slots. This has opened the door for consumers to stuff up to 128GB of memory inside a single system when all the memory slots are populated with 16GB DDR4 modules. Although quad-channel memory is the norm on the X399 platform, you can still conceivably use memory modules in a dual-channel or even single-channel configuration. However, that would defeat the purpose of acquiring a HEDT platform in the first place.

G.Skill Trident Z RGB DDR4-4000 C18 8x8GB Memory Kit: F4-4000C18Q2-64GTZR

The memory kit that we're using hails from G.Skill's Trident Z RGB product line. It is made up of eight memory modules that are produced on a ten-layer PCB (printed circuit board) with Samsung b-die chips. Each memory module is 8GB and ticks at 4,000 MHz. The timings are configured at CL18-19-19-39 with an operating voltage of 1.35V.

G.Skill's Trident Z RGB memory respects the JEDEC standard and runs at 2,133 MHz with CL15-15-15-36 timings and 1.20V operating voltage. Nevertheless, the memory modules come with two XMP 2.0 profiles. The first makes them operate at 3,733 MHz, and the second at 4,000 MHz with the advertised timings.

MSI X399 Gaming Pro Carbon AC

The MSI X399 Gaming Pro Carbon AC is the motherboard of choice for our X399 test system. The motherboard has a robust phase power delivery with a 10+3 phase design. Ten phases tend to the needs of the processor while the other three phases take care of the SoC (system-on-chip). MSI endowed the X399 Gaming Pro Carbon AC with eight DDR4 memory slots and support for memory speeds up to 3,600 MHz, and beyond. The motherboard also supports ECC (error-correcting code) memory modules.

MSI's DDR4 Boost technology has pretty much become a standard on the brand's motherboards. It's basically an optimized memory circuit design where memory slots are wired straight to the processor with no impeding components in between that can interfere or degrade the signal. As an added measure, MSI also isolates the memory circuit from the rest of the motherboard, which further helps reduce interference. Visually, it's like having a PCB within a PCB with a fence between the two.

Motherboard PCBs are produced with a combination of resin and fiberglass. In an ideal world, memory traces should be placed over the fiberglass to maintain the integrity of the memory signals. However, they are sometimes hampered by resin holes that can interfere with the signals. MSI's solution to the fiber weave problem is to distribute the memory traces in a zig-zag pattern to circumvent the resin holes.

MORE: Best Memory
MORE: DDR DRAM FAQs And Troubleshooting Guide
MORE: All Memory Content
MORE: How to Overclock a CPU

Test System and Setup

Memory manufacturers are constantly releasing new products. One of the many improvements that an updated firmware can bring to a motherboard is support for memory kits that are released after a motherboard has launched. Therefore, we updated the firmware for our MSI X399 Gaming Pro Carbon AC to the latest revision to ensure the best performance and compatibility with the latest memory kits. 

We did our testing with the BIOS settings at stock default values minus the timings for the different memory kits. For the gaming tests, we activated Game Mode in the Ryzen Master utility. Since not all games are capable of taking advantage of the Threadripper 2970WX's 24 cores, Game Mode essentially converts the 2970WX into a hexa-core processor by disabling the other cores.

On the software side, we utilized clean 64-bit installation of Windows 10 Professional with the October 2018 Update. As always, we updated the benchmarking programs and game clients to the latest public version available. Here's a thorough list of the parts in our test setup.

The scope of the article is to evaluate memory performance on the X399 platform, so we picked the most common memory speeds and corresponding timings. We tested eight different memory speeds that range from the standard JEDEC DDR4-2133 to DDR4-3333. A processor's IMC (Integrated Memory Controller) comes into play when testing high-speed memory. Unfortunately, our Rzyen Threadripper 2970WX sample hit a wall at DDR4-3333. While it was able to boot in a four-stick configuration at DDR4-3466, the system was on the borderline of instability.

While the MSI X399 Gaming Pro Carbon comes with many useful memory presets, we opted to set each memory speed manually. For good measure, we corroborated with CPU-Z that the settings are correctly applied when inside the operating system. We ran each benchmark three times and picked the median value as the final result for comparison sake. Below is a list of all our test speeds and timings. On the next page, we'll dig into the tests, and start to get a sense of how memory speeds affect performance on the X399 platform.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Office and Productivity

Adobe Creative Cloud

Each Adobe program behaves differently to memory speeds. DDR4-3333 C16 performs the best in Illustrator, Photoshop and Indesign while DDR4-3200 C14 is the best memory setting for After Effects. On the whole, memory speeds, such as DDR4-2933 C14 and higher, can provide gains of over 10% in comparison to DDR4-2133 C15.

Web Browser

Fast memory has some impact on web browsing. DDR4-3200 C14 runs almost 7% faster in MotionMark 1.1 and around 2.79% in WebXPRT 3. DDR4-3000 C14, however, is the better performer in the Kraken JavaScript benchmark with a performance gain of 3.75%.

Productivity

The difference between the faster and slowest memory speed in video conferencing and application start-up is less than 4% and 7%, respectively. The benefits of running fast memory with spreadsheet work are less than 5%. Writing workloads, on the other hand, are a big surprise here. Memory speeds above DDR4-2666 C15 provide performance boosts of over 10% with DDR4-3200 C14 performing up to 19.98% faster than DDR4-2133 C15.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Rendering, Encoding, and Compression

Rendering

Surprisingly, DDR4-2933 C14 is the fastest memory speed in Cinebench R15, being around 2.73% faster than standard DDR4-2133 C15. As for PCMark 10, DDR4-3200 C14 and DDR4-3333 C16 can accelerate photo editing, rendering, and visualization workloads by more than 11%. For video editing, DDR4-2933 C14 brings gains up to 5.98%.

Performance gains in POV-Ray and Blender are less than 2%. As a matter of fact, faster memory actually puts in worse results in the two aforementioned applications. V-Ray is indifferent to memory speeds. When it comes to the Corona 1.3 benchmark, DDR4-3200 C14 performs better than the other memory speeds and delivers a performance increase up to 8.9%.

LuxMark simply loves fast memory. Memory configurations above DDR4-2800 C14 offer gains over 30% and 20% in the OpenCL and C++ tests, respectively. What comes unexpectedly is that DDR4-2933 C14 is the best performing speed in the C++ territory with gains up to 37.42%.

Encoding and Compression

Configurations above DDR4-2800 C14 increase performance by more than 20% in compression workloads in both 7-Zip and Winrar. DDR4-3200 C14 is the absolute king of performance with gains of 27.61% in 7-Zip and 28.14% in Winrar. As expected, memory speeds don't impact decompression workloads.

Moving over to video encoding, we see noticeable benefits in running high-speed memory. DDR4-3000 C14 and faster memory modules boost encoding performance above 20% with the X264 and 13 percent with the X265 codec in HandBrake. Audio encoding isn't impacted by memory speeds.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Gaming

1280x720 Gaming Benchmarks

Memory configurations above DDR4-2933 C14 deliver performance gains of over 20% in Shadow of the Tomb Raider and F1 2018. In comparison to the first two titles, improvements in Strange Brigade are a bit lower with gains that top out at 14.1% on both the DirectX 12 and Vulkan APIs.

DDR4-2800 C14 and above offer pump out 20% more average frames per second in Assassin's Creed: Odyssey. To achieve the same 20% improvement in Far Cry 5, you need to use DDR4-3000 C14 or faster.

1920x1080 Gaming Benchmarks

At 1920x1080, Shadow of the Tomb Raider continues to see performance gains in the 20% range from memory speeds above DDR4-2933 C14. On the contrary, F1 2018 maxes out at 18.49% with DDR4-3200 C14. Benefits for Strange Brigade diminishes greatly as the improvements are now less than 3% and 2% on the DirectX 12 and Vulkan APIs, respectively.

With the increased resolution, you now need to hit DDR4-3000 C14 to see the same 20 percent gains in Assassin's Creed: Odyssey. Regarding Far Cry 5, DDR4-3200 C14 delivers 17.78% more average frames per second.

2560x1440 Gaming Benchmarks

Shadow of the Tomb Raider and F1 2018 still benefit from running high-speed memory even at the 2560x1440 resolution. DDR4-3200 C14 increases average frames per second by as much as 14.29% in the former, and around 8.63% in the latter. Memory speeds make no difference in Strange Brigade.

DDR4-3200 C14 still performs best in both Ubisoft titles. The configuration offers improvements of 7.94% and 15.56% higher average frames per second in Assassin's Creed: Odyssey and Far Cry 5, respectively.

3840x2160 Gaming Benchmarks

Shadow of the Tomb Raider and Strange Brigade show no variations in performance with memory at different speeds. Meanwhile, F1 2018 and Assassin's Creed: Odyssey exhibit gains less than 3% and 7%, respectively, while the differences in Far Cry 5 are less than 3%.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Performance Summary

General office and productivity workloads can benefit as much as 7.46% more with the DDR4-3200 C14 configuration as opposed to the standard DDR4-2133 C15 configuration.

When more specialized workloads are concerned like rendering, AMD's recommended DDR4-2933 configuration will suffice. It has the potential to boost your rendering jobs by up to 10.78%. DDR4-3200 C14 closes in a close second with a gain of 9.19%.

The DDR4-3000 C14 and DDR4-3200 C14 memory speeds are tied when it comes to encoding workloads. The former can provide you with performance gains up to 11% while the latter delivers improvements as high as 11.8%.

Moving into compression territory, we find DDR4-3200 C14 to be the best memory speed that can up your performance by 13.08%. Despite its more loose timings, DDR4-3333 C16 finishes second with an improvement of 12.45%.

Memory configurations above DDR4-2400 C14 increase average frames per second by over 10% when gaming on the 1280x720 resolution. DDR4-3200 C14 provides the largest gain with a 20.52% improvement.

You need to hit DDR4-2800 C14 to get the same 10% boost in average frames per second at 1920x1080. DDR4-3200 C14 is still the best memory speed, but the gains top out at 14.38%.

Memory speeds start to lose impact at higher resolutions. The best performing memory speed at 2560x1440 is DDR4-3200 C14, which brings in an improvement of 7.56%.

At 3840x2160 (4K) resolution, there is no point in using faster memory. The difference between the best and worst-performing configurations is less than 2%.

Memory Pricing and Conclusion

A Quick Look At Memory Prices

Having 32GB of memory sounds like overkill but is a good starting point on HEDT systems considering the nature of the workloads that are put through them. It provides enough breathing room for using many demanding applications simultaneously or when you're working with very large projects. For this exact purpose, we'll take a look at the pricing for 32GB (4x8GB) quad-channel memory kits.

In this day and age, there is no reason to pick up DDR4-2133 memory kits unless you live in a country where there is a big gap in pricing. On this side of the globe, the difference between a DDR4-2133 and DDR4-2400 is usually a couple of dollars. With the current memory pricing, you can still get decent memory kits as high as DDR4-2800 for under $200. However, if you're pursuing maximum performance, the faster memory kits like DDR4-2933 and above span from $200 to $300 depending on the brand and where you shop.

Although you might be tempted to grab the fastest memory kit, it's important to remember that the IMC inside your processor has the final word on whether the you'll be able to run the memory kit. AMD guarantees support up to DDR4-2933, so if you want the best plug-n-play experience, that's the safest memory speed to get. Running memory above AMD's specifications can require some tweaking time, or in a worst case scenario, you're forced to run the memory modules at a lower speed than the advertised. With that in mind, it's highly recommended to pick up a single kit of the capacity that you're looking for and avoid mix and matching single memory modules, which could save you money but give you a lot of headaches afterwards.

Conclusion

AMD's Ryzen Threadripper 2000-series processors perform noticeably faster in certain situations when paired with fast memory kits. Starting with office and productivity, the difference between the DDR4-2133 C15 and DDR4-3200 C14 is around 7.56%. If we look at encoding and compression workloads, DDR4-3200 C14, on average, speeds up system performance by up to 11.8% and 13.08%, respectively. The only area where we find DDR4-3200 C14 losing out to DDR4-2933 C14 is in the rendering section of our tests. Even then, the difference is less than 2%, and largely negligible in most scenarios.

If you plan to use your work system for a bit of leisure activity, having fast memory inside your rig can help increase your average framerates. If you game at 1280x720, which we hope you're not, your average frames per second can increase as much as 20.52% with DDR4-3200 C14 in comparison to DDR4-2133 C15. The performance gains at 1920x1080 and 2560x1440 are 14.38% and 7.56%, respectively. Lastly, as we've known for some time now, the impact of memory speed on 4K gaming is practically null.

In conclusion, the DDR4-3200 C14 configuration yields the most benefits overall. That's the memory speed that you should be shopping for provided that you're a performance seeker. If you're looking for something easier on the pockets, a DDR4-3000 C14 memory kit should fit that description quite well since it doesn't fall far behind the DDR4-3200 C14 standard in terms of performance.

Image Credits: Tom's Hardware

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Memory Overclocking on X470 Pinnacle Ridge

In 2018, AMD launched its second-generation Ryzen desktop processors along with the accompanying top-end X470 chipset. The current Ryzen 2000-series chips, codenamed Pinnacle Ridge, are and will continue to be popular choices for many modern PC builds, even after AMD's Ryzen 3000-series processors arrival later this year. Whether you're building a budget productivity PC, a high-end gaming rig, or a professional-level workstation, there's a Ryzen chip that will suit your needs well. But if you're planning a Ryzen build around the X470 platform, then you’re probably on the hunt for the perfect memory kit to complement your system.

As the saying goes, a system is only as fast as its slowest part. Choosing a memory kit for your system should be fairly simple, but unfortunately, it isn't. Several factors, like capacity, speed, and aesthetics come into play. Selecting the looks and capacity is pretty straightforward, but choosing the right memory speed deserves more attention.

For the scope of this article, we evaluate the benefits of using high-speed memory on AMD's X470 platform with a Pinnacle Ridge processor. We test memory speeds that start from the baseline DDR4-2133 configuration up to DDR4-3466 in various scenarios. Afterward, we examine if the performance difference (if there's any) is worth spending extra based on current memory pricing.

Pinnacle Ridge and DDR4-2933

Memory support on AMD’s Ryzen desktop processors has come a long way. The first-generation Ryzen chips, codenamed Summit Ridge, officially came with support for dual-channel DDR4 memory modules with speeds up to 2,666 MHz. With the Ryzen 2000-series processors, AMD bumped the supported speed from 2,666 MHz to 2,933 MHz, a significant increase. As we know, the majority of computer hardware can perform outside the manufacturer’s specifications as long as it does so within reasonable limits. So we expect Pinnacle Ridge processors to get along just fine with memory modules that exceed AMD’s official specification, just as memory on Intel's Z390 platform can well exceed the official 2,666 MHz limit advertised by Intel.

Fiddling around with memory timings and voltages inside a motherboard’s BIOS is certainly not everyone’s idea of a good time, and AMD knows that. For this reason, the chipmaker introduced the AMD Memory Profile (AMP) standard to facilitate the entire setup process. Basically, AMP is AMD’s own take on Intel’s proprietary Extreme Memory Profile (XMP) solution. With AMP, you can set your memory to run at its advertised speed with a couple of clicks inside an AMD motherboard's BIOS.

G.Skill Trident Z RGB DDR4-4400 C18 2x8GB Memory Kit: F4-4400C18D-16GTZR

For this article, we’re using two DDR4-4400 memory kits from G.Skill's Trident Z RGB product line. Each memory kit is 16GB and is made up of two 8GB memory modules. The memory kits run at an advertised speed of 4,400 MHz with CL18-19-19-39 timings and a 1.40V operating voltage.

G.Skill produced the DDR4-4400 memory modules on a ten-layer PCB (printed circuit board) with some of the best Samsung B-die chips. Out of the box, the memory modules operate at JEDEC’s standard DDR4-2133 with CL15-15-15-36 timings at 1.20V. However, the XMP 2.0 profile baked into the memory modules enables them to operate at 4400 MHz.

MSI X470 Gaming Pro Carbon

On the motherboard side of things, we’re using the MSI X470 Gaming Pro Carbon as the foundation for our X470 testbed. MSI has equipped this particular motherboard with a decent 8-phase power delivery subsystem. There are also four DDR4 memory slots, compatible with memory modules that run up to a speed of 3,466 MHz.

The X470 Gaming Pro Carbon motherboard employs MSI’s proprietary DDR4 Boost memory circuit design. The memory slots are connected directly to the processor, so there aren't any components that interfere with the communication between both parties. The memory traces are short by nature, so there's a lower possibility of interference as well. Additionally, the memory circuit has been isolated from other motherboard components to reduce the possibility of interference further. Essentially, this design resembles a PCB that’s housed inside another PCB, with a border that divides the two.

In the field of memory traces, there’s a phenomenon commonly known as the fiber weave effect. The motherboard's PCB is basically produced with a combination of two materials, resin, and fiberglass. More often than not, the memory traces on the PCB are obstructed by the resin holes, which in turn degrades the signal. MSI’s solution plots the memory traces on the X470 Gaming Pro Carbon motherboard in a zig-zag pattern, so the memory signals are always transported over fiberglass to avoid the dreaded resin holes.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

MORE: How to Overclock a CPU

Test System and Setup

Since our X470 Gaming Pro Carbon's firmware was a few revisions behind out of the box, we updated the motherboard to the latest public BIOS. Memory support can improve tremendously from one revision to another. It's good practice to always use the latest revision for your BIOS to ensure proper support for new memory kits. For our testing, we left all the BIOS settings at their default stock values.

For the operating system, we used a fresh 64-bit installation of Windows 10 Professional with the October 2018 Update. We also updated the test system's drivers, benchmarking programs, and game clients to the latest versions available. Here's a thorough list of the parts in our test setup.

We picked and tested eight different memory speeds starting from JEDEC's DDR4-2133 guideline to DDR4-3466. Rather than utilizing the default profiles inside the BIOS, we've opted for a manual setup. Afterward, we verified the memory speed inside the operating system with CPU-Z. We ran each benchmark three times and picked the median value as the final result for comparison between each memory configuration. Below is a list of all our test speeds and timings. On the next page, we'll dig into the tests, and start to get a sense of how memory speeds affect performance on Ryzen.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Office and Productivity

Adobe Creative Cloud

Memory speeds above DDR4-3000 perform up over five percent faster than standard DDR4-2133 in Adobe's Creative Cloud test suite. The DDR4-3200 C14 and DDR4-3200 C16 configurations provide a performance increase of 5.03 percent and 5.58 percent, respectively.

Web Browser

When it comes to web browsing, DDR4-3200 C14 is the best performer in the MotionMark 1.1 and WebXPRT3 benchmarks. It boosts performance by 3.24 percent and 3.4 percent in comparison to standard DDR4-2133. As for the Kraken JavaScript benchmark, DDR4-3000 C14 is around 1.86 percent faster than DDR4-2133.

Productivity

Memory speed has a minimal impact on video conferencing, since the difference between the best and worst result is less than two percent. Workloads such as photo editing and spreadsheets receive a performance increase of 3.87 percent and 4.06 percent, respectively with the DDR4-3466 C16 configuration. Lastly, DDR4-3200 C14 and DDR4-2800 C14 perform the best in writing and application start-up scenarios, respectively.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Rendering, Encoding, and Compression

Rendering

DDR4-3200 C14 is the fastest in Cinebench R15 and rendering and visualization workloads in PCMark 10. It delivers up to a 2.5 percent increase in the first and a 7.29 percent boost in the latter. This configuration remains the best choice in LuxMark as well. DDR4-3200 C14 is 8.45 percent and 9.16 percent faster than JEDEC's DDR4-2133 standard in the OpenCL and C++ tests, respectively.

However, DDR4-3466 C16 manages to beat DDR4-3200 C14 in the Corona 1.3 benchmark. The performance improvement of using DDR4-3466 over DDR4-2133 is around 6.41 percent.

POV-Ray is immune to faster memory as the difference between the different speeds was less than one percent. On the other hand, V-Ray and Blender show some improvement. Memory speeds in the likes of DDR4-2933, DDR4-3000, and DDR4-3200 are 2.63 percent quicker than DDR4-2133. In Blender, we see DDR4-3200 C14 once again with a 2.04 percent boost in performance.

Encoding & Compression

So far, compression workloads benefit the most from faster memory. In 7-Zip, DDR4-3200 C14 delivers up to 12 percent better performance in comparison to DDR4-2133. The same configuration performs up to 14.66 percent faster in Winrar as well. However, memory speeds play a null role in decompression in both programs.

The encoding results reveal that DDR4-3200 C14 is still the best memory speed, as the configuration delivers a performance boost of up to 3.73 percent and 2.79 percent in HandBrake with the X264 and X265 codecs, respectively.

And lastly, memory speed doesn't matter when it comes to encoding audio files.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Gaming

1280x720 Gaming Benchmarks

DDR4-3200 C14 and DDR4-3466 C16 provide the best gaming experience in Shadow of the Tomb Raider, with performance gains of 19.79 percent when compared to DDR4-2133 C15. Both configurations perform 15.52 percent quicker in Strange Brigade with the DirectX 12 API. While on the Vulkan API, DDR4-3466 C16 outperforms DDR4-2133 C15 by up to 16.48 percent. However, DDR4-3200 C14 reclaims the performance throne in F1 2018 with an improvement of 16.34 percent over baseline DDR4-2133 C15.

The back-and-forth fighting between both configurations continues with the Ubisoft titles. DDR4-3466 C16 manages to edge out DDR4-3200 C14 in Assassin's Creed Odyssey as the configuration is 15.49 percent faster than DDR4-2133. While in Far Cry 5, DDR4-3200 C14 comes out on top with an 11.96 percent improvement.

1920x1080 Gaming Benchmarks

This time around DDR4-3200 C14 delivers the best performance in Shadow of the Tomb Raider at 1920x1080 with an 18.75 percent improvement in average frames per second. For Strange Brigade, DDR4-3200 C14 is faster with the DirectX 12 API while DDR4-3466 C16 is better with the Vulkan API. The DDR4-3200 C14 and DDR4-3466 C16 configurations tie in F1 2018, as both pump out 18.62 percent higher average frames per second.

In Assassin's Creed Odyssey, DDR4-2800 C14 takes the crown with an improvement of 9.86 percent while DDR4-3200 C14 is the fastest configuration in Far Cry 5 with a gain of 12.22 percent.

2560x1440 Gaming Benchmarks

At the 2560x1440 resolution, DDR4-3200 C14 and DDR4-3466 C16 offer 10.99 percent better performance in Shadow of the Tomb Raider. But performance in Strange Brigade remains the same, regardless of memory speed and API. While in F1 2018, DDR4-3200 C14 provides up to 7.09 percent higher average frames per second.

DDR4-3466 C16 is the fastest configuration for both Assassin's Creed Odyssey and Far Cry 5. It's 7.81 percent and 17.05 percent faster than DDR4-2133 C15 in the respective titles.

3840x2160 Gaming Benchmarks

Faster memory has no impact on Shadow of the Tomb Raider and Strange Brigade at the 3840x2160 resolution. In F1 2018, the difference in performance between memory speeds is less than three percent. As for the two Ubisoft games, the difference between the fastest and slowest configuration is about 4.08 percent in Assassin's Creed Odyssey and less than two percent in Far Cry 5.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Performance Summary

Our testing shows that memory speeds don't play a significant role in general office activities and productivity. The performance difference between DDR4-2133 C15 and DDR4-3200 C14 is less than three percent. However, if you're a professional Adobe user, DDR4-3200 C14 and DDR4-3466 C16 can increase your performance by roughly five percent.

In general, DDR4-3466 C16 is the best memory speed for rendering workloads. The 3466 configuration tops the charts with a 5.03 percent gain over DDR4-2133 C15.

Like office and productivity, encoding tasks also don't benefit much from faster memory. The difference in performance is around 2.16 percent with DDR4-3200 C14.

In terms of compression workloads, DDR4-3200 C14 is the best memory speed for the job. In comparison to DDR4-2133 C15, DDR4-3200 C14 can perform up to 6.45 percent faster.

For 1280x720 gaming, memory configurations above DDR4-2666 C15 provide a performance gain of over 10 percent in the games we tested.

The impact of memory speed gains starts to decrease at resolutions above 1280x720. At 1920x1080, you need to use DDR4-3200 C14 to get the same 10 percent performance boost.

DDR4-3466 C16 and DDR4-3200 C14 deliver over 6 percent more average frames per second when gaming at 2560x1440 resolution. 

When we hit 4K (3840x2160) resolution, memory speeds practically make no difference. DDR4-3200 C14 provides a mere 1.25 percent gain over DDR4-2133 C15.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Memory Pricing and Conclusion

A Quick Look At Memory Prices

Given its volatile nature, memory pricing can change from one day to another. According to the latest trend, memory prices are falling and will probably continue to drop. Currently, there is a sizeable difference in pricing between slower and faster memory kits. Nevertheless, this gap can decrease in the upcoming months.

Examining the different 16GB (2x8GB) memory kits on the market reveals that lower-tier memory kits, like DDR4-2133 and DDR4-2400, are similarly priced. If that's the case where you live, it makes sense to pick up the faster of the two. The same can be said for DDR4-2666 and DDR4-2800 memory kits, which you can practically purchase for under $100. As expected, the pricing increases exponentially as we go up the memory speed ladder. For example, memory kits above DDR4-2933 typically start at $140.

Faster memory kits are more expensive because manufacturers bin the memory modules to run at the marketed speed without hiccups. Basically, you're paying for that plug-n-play higher-speed experience. Alternatively, you can save a few dollars by purchasing a slower memory kit and overclocking it to a higher speed. However, you'll be at the mercy of the silicon lottery draw. There's no guarantee you'll be able to achieve your desired speed. Something else to take into consideration is that memory speeds above AMD's reference DDR4-2933 are considered overclocking. Even with all the processor and motherboard advancements, sometimes a bit of manual tweaking is required to get high-speed memory kits to play nice on AMD's platform.

Conclusion

The level of impact that memory speeds have on an AMD X470-based system depends greatly on the nature of the workloads that you plan to execute. In some instances, the difference in performance can be noticeable, and at other times, it's negligible.

If you're building a daily or office system, the speed of memory shouldn't be a major concern. We saw little benefit in the majority of everyday tasks. Nevertheless, if Adobe programs are your daily bread and butter, you can boost your productivity by up to five to six percent with high-speed configurations like DDR4-3466 C16 and DDR4-3200 C14, respectively.

With more specialized workloads, memory speed can yield some benefits. If you run rendering tasks on your system, the DDR4-3466 C16 configuration is probably the best. Nevertheless, the results show that not every software behaves the same. For instance, memory speed has little to no impact on POV-Ray, V-Ray, and Blender while LuxMark, which is based on Luxrender, benefits considerably with faster memory. When it comes to encoding workloads, DDR4-3200 C14 is the way to go. However, the boost is mostly specific to encoding video files as audio encoding is pretty much immune to faster memory. On the X470 platform, compression jobs with 7-Zip or WinRar truly take advantage of memory that's operating at very high clocks. Once again, DDR4-3200 C14 is the performance king in this segment. Unfortunately, decompression gets zero benefits from running higher memory speeds.

The chances that you'll be gaming these days at 1280x720 resolution are pretty slim. If you're part of that almost extinct crowd, memory speeds above DDR4-2666 can improve your gaming experience. If you game at higher resolutions, such as 1920x1080 and 2560x1440, you can opt for configurations like DDR4-3200 C14 and DDR4-366 C16 as they can boost your average frames per second by up to 10.22 percent and 6.77 percent, respectively. At 4K (3840x2160), you don't have to worry about memory speeds as the graphics card is most likely the bottleneck.

Image Credits: Tom's Hardware

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Updated, 4/10/19, 6am PT: Adobe has officially discontinued its Shockwave product line. The URL where the Shockwave Player extensions used to be found now leads to an FAQ article about the software's demise. (Or, as Adobe put it, "End of Life of Adobe Shockwave.") The company was frank in its response to why Shockwave is going the way of the dodo, saying that "as technologies evolve and the use of mobile devices has grown, interactive content has moved to platforms such as HTML5 Canvas and Web GL and usage of Shockwave has declined." Adobe also reiterated that its enterprise customers will still receive Shockwave updates and support until the end of their current contracts.

Original article, 3/11/19, 12:36pm PT:

Adobe is putting the final nail in Shockwave’s coffin. In an FAQ article on its website, the company announced that it will discontinue the Windows version of the multimedia plugin on April 9.

This has been a long time coming. Adobe discontinued the Adobe Director app used to create Shockwave content in February 2017, and the company stopped offering the macOS version of the Shockwave player in March 2017.

While the Shockwave player for Windows will be discontinued on April 9, some customers will be able to use it until their contracts expire. Enterprise users can keep riding the Shockwave until sometime in 2022.

Shockwave is similar to Flash in that it’s used for web-based games and animations. It’s also used to make content for CD-ROMs, which at this point are essentially personifications of anachronism, thanks to digital platforms.

It's hard to begrudge Adobe for not investing in a platform like Shockwave anymore. The web has reached a point where open technologies can, and perhaps should, take the plugin’s place. And who uses CD-ROMs anymore?

Adobe is also planning to kill Flash by 2020, leaving multimedia content to open technologies instead of proprietary plugins. This is also an improvement as far as security and usability go.

That leaves the fate of content that relies on the Flash and Shockwave plugins unclear, but it should also bring an end to the days of needing to install additional software to play games or watch videos. Hopefully plugins are the next CD-ROMs.

Update 4/2/19, 7:45 a.m. PT: Last week Nvidia revealed more information about the first Creator Ready Driver release. The company said this first driver "comes optimized for a number of exciting app updates" that accelerate performance on its RTX GPUs, including Autodesk Arnold 5.3, Unreal Engine 4.22, REDCINE-X PRO 51, Adobe Photoshop Lightroom CC and Substance Designer by Adobe.

Nvidia also claimed performance improvements of up to 8-12 percent in Blender Cycles, Cinema 4D, Adobe Photoshop CC and Adobe Premiere Pro CC. Although, as always, the improvements will vary based on the rest of the system. The first Creator Ready Driver is available now via the GeForce Experience software.

Original article, 3/21/19, 8 a.m. PT:

People are used to Nvidia releasing driver updates that improve support for the latest in gaming via the Game Ready Driver program. Yesterday the company announced a similar program for creative tools called, fittingly enough, the Creator Ready Driver program.

"To achieve the highest level of reliability," Nvidia said in its announcement, "Creator Ready Drivers undergo extensive testing against multiple revisions of the top creative applications." That should improve performance in old and new versions of those apps.

The company isn't testing these apps in a vacuum. (Mostly because they can't run PhotoShop.) Nvidia said it will "conduct exhaustive multi-app testing for each type of creative workflow, evaluating driver quality in the same manner that creators work day-to-day."

Nvidia offered an example of such a workflow: using Adobe Premiere Pro CC to cut a video, sending it to After Effects CC for post-production and then kicking it back to Premiere Pro for rendering. Creator Ready Drivers focus on that process, as well as its parts.

The company was careful to note that its Creator Ready Drivers are supplementing its Game Ready Drivers, not replacing them. It said:

"Both Game Ready Drivers and Creator Ready Drivers will include the full Nvidia feature set and application support for games and creative apps, so users can continue to use either driver they prefer. But creators now have an option to receive designated driver releases with more in-depth testing to meet the stringent demands of their work."

Nvidia has released the first Creator Ready Driver; it's available via its website and the GeForce Experience app.(find it by opening the menu in the top-right corner of the window). Future releases will be "timed to key creative application updates."

Creator Ready Drivers will include support for Turing-based GeForce RTX, GTX and Titan GPUs; the Volta-based TITAN V, Pascal-based GeForce GTX and Titan GPUs and "all modern Quadro GPUs." Nvidia said they are optimized "for all the top creative applications."

Memory Overclocking on Z390 Coffee Lake

The secret to building an optimal computer system is the correct selection of its components. Careful selection not only ensures that all your components play nice together, but it also helps avoid bottlenecks that hinder performance and impact the user experience.

It goes without saying that memory is often one of the less-thought-about components. You could just purchase the most expensive memory kit, or the cheapest, but that's not always the best route. There are a plethora of memory kits on the market with different capacities, speeds, and timings. Choosing the right one can be a bit overwhelming for the first-time computer builder, or even for some of the more seasoned veterans.

Coffee Lake Refresh and DDR4-2666

Processor manufacturers, whether it be AMD or Intel, always list a processor’s supported memory speed as a guideline for the best plug-n-play experience. In the case of the 9th-generation Intel Core processors, aka Coffee Lake Refresh, Intel touts support for memory modules with speeds up to 2,666 MHz in a dual-channel configuration. Although Intel doesn't explicitly support memory speeds above DDR4-2666, the majority of Coffee Lake Refresh processors can easily handle faster memory.

Over the last couple of years, memory overclocking has become mainstream and easy. Memory manufacturers have smashed the speed limits for DDR4 memory repeatedly. Ironically, Intel itself developed the Extreme Memory Profile (XMP) standard to help consumers overclock their high-speed memory kits without much hassle.

Today we're taking a closer look at how overclocking your RAM above the recommended settings impacts performance on Intel's Z390 platform with a Coffee Lake Refresh processor, then we'll circle back and examine if that extra performance is worth it based on today's memory pricing. 

G.Skill Trident Z RGB DDR4-4400 C18 2x8GB Memory Kit: F4-4400C18D-16GTZR

G.Skill supplied a pair of Trident Z RGB DDR4-4400 memory kits for this article. Each memory kit has a capacity of 16GB and comes with two 8GB memory modules. The sticks clock in at 4,400 MHz and have the timings configured to CL18-19-19-39 with an operating voltage of 1.40V.  

The memory modules for the Trident Z RGB DDR4-4400 kit are built on a ten-layer PCB (printed circuit board) with the highest-quality hand-picked Samsung B-die chips. Upon installation, the memory modules default to JEDEC’s standard DDR4-2133 with CL15-15-15-36 timings and 1.20V. However, the sticks come equipped with a single XMP 2.0 profile. Upon activation, the profile bumps the memory modules up to DDR4-4400.

MSI MEG Z390 Ace

The MSI MEG Z390 Ace serves as the pillar of our Z390 test system. The motherboard is equipped with a robust 13-phase power delivery subsystem to house even the most demanding 9th-generation Coffee Lake Refresh processor and four DDR4 memory slots that accommodate memory modules up to 4,500 MHz.

The MEG Z390 Ace’s memory circuit design makes it an ideal motherboard for memory testing. The MEG Z390 Ace features MSI’s proprietary DDR4 Boost technology. As a result, the memory circuitry boasts an optimized and isolated design. The memory slots share a direct connection to the processor with no impeding components between them. The advantage of this design is that the memory traces are shorter, and the possibility of interference is also lower. As an added measure, MSI also isolated the memory circuit from interference originated by other surrounding components on the motherboard. MSI describes the layout like a PCB within a PCB with a canal that separates the two.

The fiber weave effect is a common problem with memory traces. A motherboard’s PCB is manufactured from resin and fiberglass, and memory traces are sometimes interrupted by the resin holes, which degrades the signal. To counteract this phenomenon, MSI routes the memory traces on the MEG Z390 Ace motherboard in a zig-zag pattern to ensure that the signals circumvent the resin holes and are constantly transported over the fiberglass instead.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Test System and Setup

We updated the MEG Z390 Ace's firmware to the latest public revision and disabled the MultiCore Enhancement / Enhanced Turbo option so that the processor abides by Intel's turbo policy. Nevertheless, MSI's motherboard runs with a 100.8 MHz base clock out of the box, which slightly overclocks the processor and memory. While this might seem like an unfair advantage in a processor review, it makes no difference in our scenario as each memory speed is overclocked by the same amount.

On the software end, we used a fresh 64-bit installation of Windows 10 Professional and installed all the available updates. We updated our test system's drivers, benchmarking programs, and game clients to the latest versions as well.

For this article, we tested 11 different memory frequencies in total that span from the standard JEDEC DDR4-2133 all the way up to DDR4-4400. Instead of applying the XMP profile, we configured the settings manually inside the BIOS. Subsequently, we verified the frequency inside the operating system with the help of CPU-Z. To ensure an acceptable level of consistency amongst the tests, we ran each benchmark three times and used the median value as the final result. 

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Office and Productivity

Adobe Creative Cloud

Overall, the DDR4-3200 C14 configuration offers the best performance in our Adobe Creative Cloud test suite. With an aggregate Adobe score of 6,740 points, DDR4-3200 was up to 15.65 percent faster than JEDEC's DDR4-2133 standard. 

Web Browser

Our web browser benchmarks yielded mixed results. DDR4-3600 C15 provided the best performance in the MotionMark 1.1 Browser test with a 7.73 percent improvement over the baseline DDR4-2133. On the other hand, DDR4-4200 C19 and DDR4-3600 C15 offered the best performance in WebXPRT 3 and Kraken, respectively. However, both were are less than 3 percent faster than the slowest configuration, so gains are slight.

Productivity

The DDR4-2800 C14 configuration came out on top in the video conferencing tests and performed roughly 4.45 percent faster than DDR4-2133. According to the test results, higher memory speeds don't have a significant impact in some scenarios, such as photo editing and application start-up: The difference between the fastest and slowest configuration was less than 2 percent and 3 percent, respectively. 

For spreadsheet work, memory speeds above DDR4-2666 actually performed worse. Curiously, DDR4-3000 delivered up to a 5.51 percent performance increase over DDR4-2133, while the other configurations showed less than 1 percent improvement.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Rendering, Encoding, and Compression

Rendering

Although the DDR4-4400 C18 configuration took the crown in Cinebench R15, it was only 2 percent faster than DDR4-2133. As for rendering and visualization workloads, DDR4-4200 C19 was the best configuration and provided a performance increase of 5.34 percent compared to DDR4-2133.

LuxMark responded very well to high-speed memory. The DDR4-4400 C18 configuration was around 86.63 percent and 13.84 percent faster than DDR4-2133 in the OpenCL and C++ tests, respectively.

DDR4-4400 continued to dominate in the Corona 1.3. There was a 12.91 percent performance improvement when using DDR4-4400 over standard DDR4-2133. Our DDR4-4400 configuration performed up to 9 percent faster in video editing, as well.

Lastly, Blender, V-Ray, and POV-Ray didn't benefit much from faster memory as the results revealed less than a 2 percent improvement.

Encoding & Compression

The overall takeaway is that high speed memory kits can greatly improve productivity in encoding and compression tasks. For a start, using DDR4-4400 memory decreased compression times in 7-Zip up to 103.4 percent when compared to DDR4-2133. WinRAR also benefited from faster memory as DDR4-4400 performed 84.92 percent faster than DDR4-2133.

In terms of decompression, neither 7-Zip or WinRAR revealed any improvements when using memory above DDR4-2133.

In comparison to DDR4-2133, the DDR4-4400 configuration improved encoding times in HandBrake up to 36.8 percent with the X264 codec and 38.14 percent with the newer X265 codec.

On the other hand, encoding audio files is fine with memory of any speed whether it be DDR4-2133 or DDR4-4400.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Gaming

1280x720 Gaming Benchmarks

In Shadow of the Tomb Raider, running DDR4-4400 C18 provided a performance increase of up to 47.90 percent in average frames per second over standard DDR4-2133. On the other hand, Strange Brigade showed a 6.57 percent improvement when using the faster memory on the DirectX 12 API and around 3.99 percent on the Vulkan API. For F1 2018, the performance improvement is around 37.85 percent when using DDR4-4400 over DDR4-2133.

Both Ubisoft titles benefited from faster memory. In Assassin's Creed Odyssey, DDR4-4400 memory increased the average frame rate by 45.71 percent over DDR4-2133. In Far Cry 5, DDR4-4400 showed a 38.79 percent improvement.

1920x1080 Gaming Benchmarks

DDR4-4200 and DDR4-4400 pumped out 33.33 percent higher average frames per second in comparison to DDR4-2133 in Shadow of the Tomb Raider at a resolution of 1920x1080. Higher memory speeds made no difference in Strange Brigade on the DirectX 12 API, and the difference with the Vulkan API is negligible. DDR4-4200 puts out 19.88 percent higher average frames than DDR4-2133 in F1 2018.

DDR4-4200 performed the best in Assassin's Creed Odyssey delivering up to 33.33 percent over baseline DDR4-2133. As for Far Cry 5, DDR4-4000 was 31.90 percent better.

2560x1440 Gaming Benchmarks

Improvement in Shadow of the Tomb Raider was less than 4 percent, and the difference in Strange Brigade is below 1 percent. The fastest memory configuration from the slowest showed a 6.58 percent difference in F1 2018.

The DDR4-4200 and DDR4-4400 configurations performed 21.88 percent better in Assassin's Creed Odyssey. As for Far Cry 5, DDR4-4000 and DDR4-4400 both delivered 19.30 percent higher average frames per second.

3840x2160 Gaming Benchmarks

Faster memory makes very little difference in gaming at the 4K (3840x2160) resolution. Titles like Shadow of the Tomb Raider, Strange Brigade, and F1 2018 revealed negligible benefits from running higher-speed memory kits. As a matter of fact, some of the faster memory kits actually performed worse than their slower counterparts.

A few configurations provided better performance in Assassin's Creed Odyssey. DDR4-3200, DDR4-3600, DDR4-4000, and DDR4-4200 increased the average frame rate up to 8 percent. In Far Cry 5, memory above DDR4-3200 performed around 2.63 percent better.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Performance Summary

DDR4-3000 C14 turned out to be the best configuration for office and productivity work. It performed 3.64 percent faster than the standard DDR4-2133. DDR4-4400 C18 came in a close second with a 3.54 percent improvement over the baseline configuration, but the top four entries are close enough to call a wash.

Generally, higher memory speeds provided better performance in rendering tasks. Memory speeds above DDR4-3600 C15 delivered over a 10 percent performance increase.

Encoding workloads benefited significantly from high-speed memory. Memory speeds over DDR4-2400 C14 provided performance bumps of over 10 percent. The DDR4-4400 C18 configuration is up to 23.64 percent faster than DDR4-2133 C15.

When it comes to compression, any memory speed over DDR4-2400 C14 contributed to a healthy performance gain of over 10 percent. The higher configurations, such as DDR4-4200 and DDR4-4400, dominated the charts with performance increases of over 30 percent in comparison to DDR4-2133 C15.

Configurations higher than DDR4-2400 C14 presented performance gains of over 10 percent during gaming tests at the 1280x720 resolution.

For gaming at the 1920x1080 resolution, DDR4-4400 C18 was the fastest configuration. However, it only provided a performance gain of 18.92 percent over DDR4-2133 C15.

Performance gains began to diminish as we scaled the resolution ladder. At 2560x1440, memory speeds above DDR4-2133 C15 only performed around 2 to 9 percent faster.

Faster memory had very little impact on high-level gaming at the 4K (3840x2160) resolution. The difference between the fastest and slowest configuration was less than 3 percent.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Memory Pricing and Conclusion

A Quick Look At Memory Prices

At the lower end of the spectrum, the difference in pricing between memory kits is minimal. You should just avoid DDR4-2133 in general as most of the time you can probably find a similarly-priced and specced DDR4-2400 kit. Scaling up to DDR4-3000 means spending an extra $60 on memory for your build budget, but it's still easy on the pockets. Once you break the DDR4-3000 mark, memory prices rise considerably.

The more expensive memory kits are binned to run at the advertised frequency, so basically you're paying for a plug-n-play feature assuming your processor and motherboard are up to the task. If you're feeling adventurous, purchasing a lower speed kit and manually overclocking it to the desired speed is an option, and you could save some coin in the process. However, it's like playing the lottery as it's impossible to assess a memory kit's overclocking potential before taking it out of the box. Provided you buy from a reputable brand, roughly half of the garden-variety memory kits should perform faster than their rated speeds.

Conclusion

After countless hours of testing, one thing is utterly clear: There isn't a single perfect memory speed for all applications. Enthusiasts ready to build a new system around one of Intel's 9th Gen, Coffee Lake Refresh processors should first decide what they'll use the system for, and then pick the most adequate memory speed accordingly.

For the average computer user that engages in everyday tasks, memory speed doesn't hold as much importance on the Z390 platform. The aggregate difference between the fastest configuration (DDR4-3000 C14) and the slowest configuration (DDR4-2133 C15) in office applications is less than 4 percent. However, if you're a professional that constantly works with the Adobe suite of programs, you could benefit from running DDR4-3200 C14 memory as it delivers a performance gain of around 15.64 percent over the baseline DDR4-2133 C15 configuration.

It's a little more complicated in regard to rendering workloads. In general, memory speeds over DDR4-3600 C15 provide a 10 percent or more performance improvement over DDR4-2133 C15. However, not all rendering software is equal, so some programs don't take full advantage of high-speed memory. It's important to identify the programs you use the most and see whether it's worth investing extra in faster memory.

If you plan to do lots of encoding work on your Coffee Lake-powered system, you'll enjoy big benefits from memory clocked at very high speeds. Testing shows that memory speeds above DDR4-3400 C16 perform over 20 percent faster than DDR4-2133 C15, which is very significant. Here's where you have to evaluate how much encoding tasks you do and whether the extra cost and tuning is justified.

Compression workloads also profit from faster memory. Memory modules that run above DDR4-3000 C14 can easily help improve your system's performance by over 20 percent. This is one of the situations where ultra-high-speed memory shines – a highly-tuned setup like DDR4-4200 C19 is up to 39.14 percent faster than DDR4-2133 C15.

If you're a gamer, memory speeds can mean either everything or nothing for you. If you still game at 1280 x 720, memory speeds higher than DDR4-2400 C14 can easily net you a performance gain over 12 percent assuming your graphics card isn't bottlenecking your system. For 1920 x 1080 gaming, the improvement isn't as much – you'll need to hit DDR4-3000 C14 to get the same 10 percent gain. Memory speeds aren't a priority at the higher resolutions. At 2560 x 1440, the difference between the fastest configuration and slowest is less than 9 percent, and at 4K (3840 x 2160), the difference is negligible. That's because you're bumping up against a graphics card bottleneck, as opposed to a limitation of the memory subsystem.

Ultimately the best choice is to build a system based upon your individual use-case, but it's important to remember that faster (and more expensive) memory isn't always the best choice.

Want to comment on this story? Let us know what you think in the Tom's Hardware Forums.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Intel must be starting to feel like Frito-Lay. It had an entire market to itself for a while, but now that other companies have started to sense weakness, it's hard to find someone that isn't planning to make its own chips. The latest would-be competitor? Well, according to Axios, it's Adobe.

That's right: the company that makes the software that creative professionals have to sell their organs to pay for each month is reportedly thinking of making its own processors. Or at least licensing Arm's designs so it can better integrate the software it's known for with the hardware on which it runs.

Adobe certainly isn't the first company to consider making--or actually make--its own chips. Axios noted that Apple, Google, Samsung, and Amazon already do just that. (And speculation runs rampant among the Apple community about if or when the company will decide to ditch Intel for good.)

Those companies don't make their own chips for the fun of it. They do it because it gives them more control over their products, rather than forcing them to make their software for standard hardware. The idea is that this leads to better performance while also reducing dependence on outside companies. Imagine that right now every product is like a flavor dust applied to a Lay’s chip. Eventually, someone was going to make their own spuds from scratch.

From that perspective, Adobe making its own chips would make sense. Its software is an ecosystem unto itself—there are people out there whose livelihoods are directly affected by their proficiency with and performance in Adobe’s creative tools. (Sorry, sorry, the Adobe Creative Cloud subscription service. Branding!) Improving performance with custom silicon would help those people and, of course, give Adobe yet another way to make itself all-but-indispensable to creators.

But there are a lot of unknown factors here. Companies don't just license Arm tech and release products with custom chips overnight. They have to hire people who know how to design processors, build the chips, optimize the software for those chips, and then follow a thousand other steps on a process we’re already oversimplifying. Does Adobe plan to hire those people? Could it win in a bidding war against, well, any of the other companies we've mentioned so far in this article?

Then there's the matter of selling the chips. Facebook, Google, and Amazon primarily use their own chips in their data centers. Even though their usage affects hundreds of millions of people, then, the vast majority of them don't care. Meanwhile, Apple and Samsung use their chips to sell their products. Those are very different scenarios even if they both start with custom silicon. The former is suited to servers that only have to support a predefined software suite. The latter has to support countless apps, utilities, and other products over which the companies have relatively little control.

It's not clear what approach Adobe might attempt to emulate in that sense. Would it make a bunch of custom chips to use in a server and then off-load compute-intensive tasks to the cloud? Or would it try to convince people to buy a processor specifically because they need Photoshop to run faster? The former is reasonable; the latter could prove to be a hard sell even for Adobe.

Best not to fret too much about it. Axios quoted Adobe CTO Abhay Parasnis as saying: “Do we need to become an ARM licensee? I don't have the answer, but it is something we are going to have to pay attention to." Parasnis was bullish on Arm—the company changed its branding a few years back, folks, it's time to accept it—but not clear on what that meant for Adobe. Probably the responses to articles like this one will help inform the company's decision.

As for us? We’re just the taste testers. You put a chip in front of us and we’re going to eat it...erm, uh, run it through a rigorous series of tests to determine how it compares to its counterparts. Definitely the second one. Tom’s Hardware has not consumed any processors. (To our knowledge.)

Each week we tap into the collective knowledge of the Tom's Hardware community to find the answers to our most pressing questions. This week we want to know how you would build a powerhouse Adobe Premiere PC.

It’s no secret that we love video at Tom’s Hardware. You may have noticed an increasing number of videos published to the site over the past few years. Whether it’s an in depth dive into our RGBeast Build or the thousands of tutorials we feature in our community, video content is a crucial part of our workflow and creative process.

With that in mind, we wanted to find out what it takes to build an affordable (that is less than roughly $2,000/£2000), but professional, video editing rig. While we are posing this question to the community, I also wanted to take a stab at the parts list to see what I could come up with.

Admittedly, it’s been a while since I had to spec out a video editing build. We produce almost all of our video content using Adobe Premiere, and so I wanted to center my parts list around this program. In my research, I was surprised to find that video editing is no longer entirely reliant on the CPU, and requires a pretty hefty graphics card these days. Storage is another prime consideration in this build. I needed to balance the needs of storage size and storage speed.

At Tom’s Hardware, we strongly encourage our users to only purchase SSDs, and this build is no exception. While the space may be limited in comparison to a hard disk based storage system, most files these days are stored in the cloud. Two terabytes should be enough to handle a handful of projects at a time, with the rest archived online. With all of that said I’m pretty happy with this 4K capable Premiere build.

Threadripper proves to be a tremendous value for workstation tasks. Just keep in mind that with great power comes great wattage, and a high TDP, which is why I went with the NZXT Kraken X62 AIO CPU cooler. Up next was the RAM, and 64 GB of quad channel memory should serve us well on the X399 platform. Tying it all together is the power supply, which is an impressively nimble 650 watt 80+ Platinum Certified PSU from Seasonic.

I’ve shown you mine, now let’s see yours. How would you change this build? Is anything missing? Let us know in the comments below.

Cover image credit: Gorodenkoff / Shutterstock

If there is one constant in the universe, it's that Adobe has to fix something with its Flash Player on a regular basis. It turns out the holiday season is no exception because earlier this week the company released another batch of security updates for the Adobe Flash Player across multiple operating systems.

Adobe said these updates are meant for Adobe Flash Player versions up to 31.0.0.148 for Windows, macOS, Linux, and Chrome OS. (So essentially all of the platforms that currently support Flash.) It didn't offer more info about the patch, but it did say exploiting the vulnerability could lead to "successful exploitation could lead to arbitrary code execution in the context of the current user," and that details are "publicly available."

PLACEHOLDER_3#

Which, of course, means that hackers already know about the vulnerability and will probably race to exploit it before people update. Because it's a holiday weekend in the U.S. and few people think about Flash as it is, that means the number of people affected by the issue could be high. Shout-out to all the IT workers spending their days updating Adobe Flash Player instead of celebrating the holidays with their loved ones.

By now Flash having security problems has become a meme. Every few months we cover a new update, whether it's Adobe and Microsoft teaming up to resolve a vulnerability in February or Adobe doing it solo in June, and we don't even write about every security problem in the utility. Even as Flash becomes irrelevant, the tool's nigh ubiquity makes it a compelling target for hackers looking to compromise as many systems as possible.

Earlier today we reminded everyone that Thanksgiving is about, well, giving thanks. At the time we said we were thankful not to be working at Microsoft because of the Windows 10 October 2018 Update's apparent inability to reach everyone in its stated launch window. Now we have something else to be thankful for: the knowledge that Flash is finally going to stop being such an attractive target when it's shut down in 2020.

RGB SSDs Done Right, Kinda

The SSD market is riding the wave of a new RGB trend, and Team Group's T-Force Delta RGB SSD is yet another example of the newfound brightness. Recently we reviewed Kingston’s HyperX Fury RGB, and while it looked like a beast, it was timid in our testing. It's clear that RGB doesn’t always mean performance–it really is more about form over function for some products.

Team Group’s T-Force Delta RGB tries to bring both of those aspects to the table. The Delta RGB reaches peak speeds of 560/510MB/s courtesy of Micron’s last-gen 32L 3D TLC NAND flash and a Silicon Motion controller. While it isn’t a leaderboard king with its mainstream performance, the Team Group’s T-Force Delta RGB is a good blend of SSD and a blingy light show. However, the somewhat low endurance, a three-year warranty, and a slight premium for the flashy lights might be deal breakers.

Team Group hit the market early with its RGB SSD. Focusing on multiple products to cover this ever-expanding gamer-focused market, Team Force developed multiple versions of their Delta RGB SSD.

You can decide between a black or white casing, but there are more options: There are different Delta RGB SSD models with different types of RGB support. It is important that you check the motherboard compatibility lists on Team Group’s website to ensure your system is compatible. Otherwise, you might be stuck with the wrong drive for your application.

We're testing the Delta RGB with 3-pin 5V ADD header support. Team Group includes a micro USB cable that connects to the SSD’s USB input and the appropriate header of your motherboard. The SSD is the first to feature a water flow light effect, which is actually the default setting. The water flow light effect cycles through the color range until you change the lighting settings with any of the following software per your respective motherboard supplier: ASUS Aura Sync, GIGABYTE RGB Fusion, MSI Mystic Light Sync, or ASROCK Polychrome Sync.

Team Group also has the Delta S, which is compatible with 4-pin 12V RGB headers, and the Delta R, which plugs into a USB 2.0 header. The Delta R, however, has only one lighting mode and cannot be controlled or synced with software. If you purchase a Delta RGB 5V variant and don’t have a motherboard to support it, you can use a Corsair Lighting Node PRO and its LINK software instead, just like we did. Most other RGB lighting controllers do not support 5V.

Specifications

Just as any other SATA SSD, sequential performance isn’t anything to write home about. The Delta RGB is capable of sequential speeds up to 560/510MB/s read/write, and random performance spans up to 90,000/80,000 IOPS read/write. Much of the write performance comes from the integrated SLC buffer.

Team Group’s Delta RGB is available in capacities of 250GB, 500GB, and 1TB. There is a slight premium for the bright and unique RGB look.

Like the Kingston HyperX Fury RGB and WD Blue 3D, the Delta RGB comes with a three-year warranty, but it doesn't have the endurance to match. Competing SSDs offer 360 to 600TBW endurance ratings at the 1TB capacity, but the Delta RGB comes with just 240TBW of endurance.

A Closer Look

Team Group’s Delta RGB comes in a 2.5” 9.5mm form factor, so it is a little thicker than the typical 7mm thick SSD. It is also slightly heavier at over 80 grams (rather than the average 30-40 grams), so it wouldn’t be a good choice as a replacement drive for a thin laptop. But that doesn’t matter much: This SSD is designed to be on display in desktops. With the extra 2.5mm of space, Team Group has added in RGB lighting and a light diffuser which results in some beautiful lighting and transitions.

Unlike the HyperX Fury RGB and its 75 embedded lights, the T-Force Delta uses just eight LEDs to produce its light range, but it can still deliver over 16.8 million colors. The Delta's LEDs reside on a separate circuit board that's powered by a separate cable, rather than being powered directly by the SSD’s PCB.

The Delta features a Silicon Motion SM2258 controller. Well established and mature, this 4-channel controller provides solid performance and uses a DDR3 DRAM cache. Pairing it with Micron’s last-gen 384Gb 32L 3D TLC, however, might not be the best combination for competitive performance, but it does provide more spare area for the SLC cache, garbage collection, and other background tasks.

The 250GB model actually has 288GB of raw NAND, with usable capacity weighing in at 238GB after formatting in Windows. We weren’t able to decode the NAND packages on the 1TB model, but it has 931GB of usable capacity after formatting. 

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

250GB Performance Results

Comparison Products

We included the Intel 905P and a 2TB WD Blue HDD in our performance charts. We also have the Crucial MX500 and Plextor M8V that come with the same SM2558 controller. The MX500 features Micron’s latest 64-layer (64L) TLC flash while the M8V hosts Toshiba’s latest 64L TLC flash. Also, we have the WD Blue 3D and Toshiba OCZ TR200 with Toshiba 64L TLC flash, and we threw in Samsung’s 860 EVO, which comes with its own in-house controller and 64L 3D TLC. Finally, we added in the Intel SSD 660p, which is one of the least expensive NVMe SSDs.

Trace Testing – PCMark 8 Storage Test 2.0

PCMark 8 is a trace-based benchmark that uses Microsoft Office, Adobe Creative Suite, World of Warcraft, and Battlefield 3 to measure the performance of storage devices in real-world scenarios.

Team Group’s Delta RGB lands in fifth place behind the Crucial MX500. Thanks to its four-channel controller and DRAM, it leads the Toshiba OCZ TR200, which only has a two-channel controller and no DRAM. If you're searching for the best performance in these types of applications, the Samsung 860 EVO is your pick.

Game Scene Loading - Final Fantasy XIV

The Final Fantasy XIV StormBlood benchmark is a free real-world game benchmark that easily and accurately compares game load times without the inaccuracy of using a stopwatch.

There isn't much difference between the SSDs during our game loading test in Final Fantasy. Most of the SATA SSDs take about 22-23 seconds to load the game, including the Delta RGB. Here the Delta RGB takes fourth place. It is also much faster than the Toshiba OCZ TR200.    

Transfer Rates – DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with our own custom 50GB block of data. Our data set includes 31,227 files of various types, like pictures, PDFs, and videos. We copy the files to a new folder and then follow up with a read test of a newly-written 6 GB file.

The Delta RGB delivers average file copying performance. The Samsung 860 EVO takes the lead when copying and reading data, but the Delta RGB nearly matches the M8V, which also has an SM2258 controller. It even outpaces both the WD Blue 3D and Toshiba OCZ TR200. It's almost three times faster than an HDD during the file copy test and over four times faster during the read test.

SYSmark 2014 SE

Like PCMark, SYSmark uses real applications to measure system performance. SYSmark takes things much further, however. It utilizes fourteen different applications to run real workloads with real data sets to measure how overall system performance impacts the user experience. BAPCo's SYSmark 2014 SE installs a full suite of applications for its tests, which includes Microsoft Office, Google Chrome, Corel WinZip, several Adobe software applications, and GIMP. That also makes it a great test to measure the amount of time it takes to install widely-used programs after you install a fresh operating system.

SYSmark installed very quickly on the Team Group Delta RGB. It outperformed the HDD and landed just behind the Crucial MX500 and Intel 905P. During the performance portion of the test, the Delta RGB scored a respectable 1555 points. This lands just behind the Plextor M8V, but clearly ahead of the HDD and TR200.      

ATTO

ATTO is a simple and free application that SSD vendors commonly use to assign sequential performance specifications to their products. It also gives us insight into how the device handles different file sizes.

The Delta RGB delivers average performance during the read and write tests, although it lands on the lower side of the average. Our 250GB sample attained over 560MB/s of read throughput and over 520MB/s of write bandwidth, which should satisfy anyone looking for a new SATA SSD.

Anvil's Storage Utilities

Anvil's Storage Utility is a commonly-referenced benchmark that simplifies the complex IOMETER benchmark and its underlying Dynamo engine with a one-click software wrapper.

The newer 64L TLC flash has much faster write performance than the older 32L TLC NAND flash in the Delta RGB. While Team Group’s T-Force RGB Delta landed in third place, its low write performance isn't quite as impressive.

CrystalDiskMark

CrystalDiskMark (CDM) is a simple and easy to use file size benchmarking tool.

The Samsung 860 EVO, WD Blue 3D, and the Team Group Delta RGB offer similar sequential read and write performance. With results of 564MB/s read and 534MB/s write, there’s nothing to complain about.

Low queue depth performance is critical for SSDs because it directly impacts the user experience. At a queue depth (QD) of 1, sequential read and write speeds of 493/470MB/s are average results, but QD1 4K random read performance of 29MB/s is the slowest result in the group. 4K random write performance is also average at QD1. Once we dive into the deeper QDs of 2, 4, and 8, the Delta RGB continues to show weaker than average read and average write performance.

Sustained Sequential Write Performance

Official write specifications are only part of the performance picture. Most SSD makers implement an SLC cache buffer, which is a fast area of SLC-programmed flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the SLC cache and into the "native" TLC or QLC flash. We hammer the SSDs with sequential writes for 15 minutes to measure both the size of the SLC buffer and performance after the buffer is saturated.

The Delta RGB’s write cache is much larger than the cache on the other drives. While the Samsung 860 EVO can hold about 12GB of writes before the cache is full, the Delta RGB absorbed 90GB of data before it filled. Once the SLC write cache filled, however, the Delta wrote at an average of 90MB/s for an additional 22GB. Once again, performance fell even further to an average of 64MB/s, which is much lower than the competition.

Power Consumption

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is a very important aspect to consider, especially if you're looking for a new drive for your laptop. Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state faster, which ultimately saves power.

Team Group’s Delta RGB SSD is quite efficient. It drew an average of 1.51W during the 50GB file transfer and hit a maximum of 2.6W. With this and an average transfer speed of 120MB/s, it ranks as the most efficient 250GB SSD in its class during this workload.

At idle, the 250GB Team Group Delta RGB continues to show off its efficiency. It consumed nearly half a watt with LPM disabled, which is 2.5x more than the Samsung 860. With LPM enabled, the Delta RGB consumed under 50mW, which is a very good result.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

1TB Performance Results

Trace Testing – PCMark 8 Storage Test 2.0

PCMark 8 is a trace-based benchmark that uses Microsoft Office, Adobe Creative Suite, World of Warcraft, and Battlefield 3 to measure the performance of storage devices in real-world scenarios.

The 1TB Delta delivers a notable improvement over the smaller 250GB model, but it isn't enough to take the crown. Instead, the drive ranks 5th overall, closely behind Crucial’s MX500. 

Game Scene Loading - Final Fantasy XIV

The Final Fantasy XIV StormBlood benchmark is a free real-world game benchmark that easily and accurately compares game load times without the inaccuracy of using a stopwatch.

The 1TB Delta RGB was actually loads game scenes slower than the 250GB model. Here it loaded the scenes in 23.1 seconds, which ranks ahead of both the TR200 and WD Blue 3D.

Transfer Rates – DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with our own custom 50GB block of data. Our data set includes 31,227 files of various types, like pictures, PDFs, and videos. We copy the files to a new folder and then follow up with a read test of a newly-written 6 GB file.

The NVMe-based Intel SSD 660p teases its SATA competitors with an average of 291MB/s. But the Delta’s results are very similar to the Crucial MX500 and WD Blue 3D. The Delta RGB performs three times faster than an HDD and almost twice as fast as Toshiba’s OCZ TR200 during our file copy test.

SYSmark 2014 SE

Like PCMark, SYSmark uses real applications to measure system performance. SYSmark takes things much further, however. It utilizes fourteen different applications to run real workloads with real data sets to measure how overall system performance impacts the user experience.

BAPCo's SYSmark 2014 SE installs a full suite of applications for its tests, which includes Microsoft Office, Google Chrome, Corel WinZip, several Adobe software applications, and GIMP. That also makes it a great test to measure the amount of time it takes to install widely-used programs after you install a fresh operating system.

The Team Group T-Force Delta RGB installs SYSmark faster than any other SATA device in our comparison pool. This lead is short-lived, however. The Delta RGB ranks sixth place in the responsiveness score, just a few points below the Samsung 860 EVO and WD Blue.

ATTO

ATTO is a simple and free application that SSD vendors commonly use to assign sequential performance specifications to their products. It also gives us insight into how the device handles different file sizes.

Like the 250GB model, the 1TB model lands the lower portion of the chart but still delivers speeds of 560/525MB/s read/write. While a bit more than just a few bucks more at the 1TB capacity, the Intel SSD 660p clearly takes the lead when it comes to sequential workloads–it's over three times faster.

Anvil's Storage Utilities

Anvil's Storage Utility is a commonly-referenced benchmark that simplifies the complex IOMETER benchmark and its underlying Dynamo engine with a one-click software wrapper.

The Delta takes fifth place due to its lower than average read and write scores. We really see the weakness of the TR200's DRAMless architecture in the write performance results, but that type of workload is the Delta RGB's strength.

CrystalDiskMark

CrystalDiskMark (CDM) is a simple and easy to use file size benchmarking tool.

The Intel 660p takes the lead in sequential performance at QD32. The Delta RGB, however, delivers a respectable result of 562/526MB/s read/write. Unlike the 250GB capacity, the 1TB model delivered similar 4K results at queue depth 32 compared to the rest of the drives.

The Delta scores an average result at QD1, but, like in Anvil, the Delta is more competitive during write workloads than it is during reading tasks.

Sustained Sequential Write Performance

Official write specifications are only part of the performance picture. Most SSD makers implement an SLC cache buffer, which is a fast area of SLC-programmed flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the SLC cache and into the "native" TLC or QLC flash. We hammer the SSDs with sequential writes for 15 minutes to measure both the size of the SLC buffer and performance after the buffer is saturated.

The 660p offers three to four times the write performance but slows down to just over 100MB/s after 135GB of data writes. Like the 250GB model, the 1TB Team Group Delta RGB slows down after being written to for more than a few seconds. After writing 11GB of data at a rate of 525MB/s, our 1TB sample’s write performance degraded to 470MB/s for the remainder of the test. Compared to the Samsung, WD, and Crucial comparison products, it’s a small difference.

Power Consumption

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is a very important aspect to consider, especially if you're looking for a new drive for your laptop. Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state faster, which ultimately saves power.

We were surprised to see how closely the 1TB results were to the 250GB model. With an average of 1.55W, along with a maximum of 2.68W recorded during the 50GB copy test, the Delta RGB has very low power requirements under load. The Delta is also very efficient at 97MB/s per watt. The Samsung 860 EVO’s 108MB/s per watt and the NVMe Intel 660p’s 120MB/s per watt are better, though.

The drive sipped 0.366W of power at idle with LPM disabled. But while we were hoping to see an improvement with LPM enabled, unlike other SSDs we've tested, the Delta RGB wouldn’t drop into a lower power mode. Thus, its 0.366W lands in last place.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

Conclusion

RGB strikes, but this time around, slower performance does not totally overshadow the allure of the shiny bright lights. Unlike the HyperX Fury RGB, which we suggested for those of you chasing aesthetics above all, there isn’t as much of a compromise in performance for looks–now it is mainly just price.

Both the 250GB and 1TB capacities performed well during testing. The drives often ranked in the middle or near the tail end of the comparison pool but were still within a few MB/s or points of the MX500 or WD Blue. It can’t hold a candle to the Samsung 860 EVO’s performance, nor the Intel SSD 660p's sheer raw read and write speed, but the Delta RGB SSD will still get the job done in a timely manner with regular applications. Most importantly, unlike the DRAMless Toshiba OCZ TR200 and the Kingston HyperX Fury RGB, there's no period of laggy performance after filling the drive, bombarding it with our benchmark routine, and continuing to use it as an OS volume.

The Delta RGB's main detraction stems from its premium pricing. Both our Crucial and Samsung comparison products both offer five-year warranties while the Delta RGB only has a three-year warranty. We expect this from an entry-level SSD, but the Delta RGB's pricing isn’t in that class.

Then there's the NVMe Intel SSD 660p and ADATA XPG SX8200 that also offer five-year warranties and much more performance than the Delta RGB, but for just a few dollars more. While they can’t necessarily compete on aesthetics, they are a much better value for those looking for the most bang for the buck, especially at the 500GB capacity point.

Finally, while the Delta's endurance rating is greater than the Intel SSD 660p’s, most SSDs still offer more. This mostly is Team Group's decision to use Micron’s last-gen 32L TLC flash. But, in all, this last point is rather moot because most SSDs won’t see tens of terabytes written within the three year warranty period.

Team Group’s T-Force Delta will provide a cool aesthetic for any build, new or old. While it does come at a premium and trades off a little bit of performance here, and comes with a slightly low three-year warranty, we recommend it to those looking for an RGB SSD. It gets the job done, and it looks good doing it.

MORE: Best SSDs

MORE: How We Test HDDs And SSDs

MORE: All SSD Content

Cheap Chips

Building a capable PC for less than $500 is challenging, especially as memory prices soar. Fortunately, a competitive CPU market means it's possible to get lots of processing power without breaking the bank.

Fast, affordable Ryzen models from AMD forced Intel to improve its dual-core Pentium chips with Hyper-Threading Technology and larger L3 caches. As a result, today's Pentiums look a lot like yesterday's Core i3s. That makes them a big win for budget-oriented builders, particularly when they're paired up with Intel's H370, H310, and B360 chipsets.

Intel's victory isn't assured, though. AMD has a fearsome competitor in its overclockable Raven Ridge-based Ryzen 3 2200G, which sells for $100 and sports four execution cores plus the impressive Radeon Vega integrated graphics engine. It's quick enough for low-resolution gaming, potentially saving lots of money on a discrete GPU.

The UHD Graphics 630 solution built-into Intel's Pentium Gold G5600 can't even come close to AMD's Radeon Vega, and the G5400's UHD Graphics 610 is even slower. So, we matched the host processors up to an add-in graphics card for a more direct comparison in our benchmark suite. In the end, we found that Ryzen 3 2200G tells a better value story than the Pentium Gold G5600, while the Pentium Gold G5400 is simply unmatched at its $64 price point. 

Pentium Gold G5600 and G5400

Last year, Intel announced it was rebranding the Pentium family. Pentium Gold comprised the higher-performance socketed models based on the Kaby Lake architecture (and now Coffee Lake), while Pentium Silver CPUs were power-optimized and BGA-attached, leveraging Intel's Goldmont Plus design.

Intel also added Hyper-Threading Technology to its Kaby Lake-based Pentiums in an effort to stave off the then-impending Ryzen onslaught. That practice continues with today's Coffee Lake-based models, allowing the dual-core Pentium Gold G5600 and G5400 CPUs to operate on four threads concurrently.

Of course, Coffee Lake is manufactured using an optimized 14nm++ process. That, plus a 3W-higher thermal design power, is responsible for the 200 MHz speed-up available across the Pentium Gold family. Intel also bumped L3 cache capacity up to 4MB, a 33% increase compared to Kaby Lake-based Pentiums. The dual-channel DDR4 memory controller is still limited to 2400 MT/s, so peak bandwidth does not change. And whereas AMD's Ryzen 3 2200G only gives you eight lanes of PCIe 3.0 for discrete graphics upgrades, both Pentium chips offer a full 16-lane link. 

Pentium Gold CPUs don't get all of Intel's special sauce, though. Similar to the Core i3 models, Pentiums lack Turbo Boost functionality altogether. Under load, you get one static frequency, regardless of how many cores are active. Intel also locks its ratio multipliers, preventing overclockers from coaxing extra performance from the chips. Pentium processors don't support the AVX/AVX2 instructions that accelerate certain productivity workloads, either. As a result, AMD's AVX-enabled Ryzen 3 2200G enjoys a performance advantage in several optimized applications, as you'll see in our benchmarks. Finally, Optane memory isn't an option in Pentium-based PCs. 

Intel's Pentium Gold G5600 includes on-die UHD Graphics 630, while the G5400 utilizes UHD Graphics 610. The former is composed of 24 execution units in what is referred to as a GT2 configuration, while the latter consists of 12 EUs in a GT1 setup. A 350 MHz base graphics frequency boosts up to 1.1 GHz on the Pentium Gold G5600 and 1.05 GHz on the G5400.

UHD Graphics 630/610 supports a wide range of codecs and provides hardware acceleration for most media consumption tasks. It's also equipped with plenty of connectivity options, including native support for DisplayPort 1.2a and HDMI 1.4. But the UHD Graphics engine isn't really suitable for gaming, even at low resolutions and relaxed quality settings. Plan on adding a discrete graphics card if you plan to build a gaming PC around Intel's Pentium Gold.

Fortunately, you should have some room left in your budget for an upgrade. The G5400 model sells for a mere $64, placing it well under AMD's low-end Ryzen options. We don't expect it to face any real competition at that price point. Meanwhile, the Pentium Gold G5600 should be available at $86. But as we saw with Intel's previous-gen G4620, street pricing is much higher. You'll currently find it around $95, placing it close to AMD's Ryzen 3 2200G. Consequently, the Pentium grapples with an overclockable competitor armed with four physical cores and impressive Radeon Vega integrated graphics.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Test Setup

Test Notes

We paired our Pentium Gold processors with a Z370-based motherboard, which does support faster memory. Sticking with Intel's official specification, we used our modules at 2400 MT/s to represent the data rates you'll have access to with B- and H-series platforms.

Comparison Processors

Test Systems

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

VRMark, 3DMark & AotS: Escalation

VRMark & 3DMark

We aren't big fans of using synthetic benchmarks to measure game performance because the differences in their results often aren't representative of real games. But 3DMark's DX11 and DX12 CPU tests provide useful insight into the amount of horsepower available to modern engines.

The DX11 and DX12 tests found Intel's newest Pentium Gold models serving up nice gains over last generation's Pentiums. But Ryzen 3 2200G performed even better in stock trim, and overclocking only widened its advantage. Unfortunately for Intel, Pentiums cannot be tuned.

UL's VRMark test lets you gauge your system's suitability for use with the HTC Vive or Oculus Rift, even if you don't currently own an HMD. While every sample in our pool delivered a passing score (above 109 FPS), AMD's Ryzen 3 2200G again slipped past its Pentium competition.

Ashes of the Singularity: Escalation

Ashes of the Singularity is a heavily-threaded game that runs best on CPUs with lots of cores. Ryzen 3 2200G technically led the pack of comparably-priced processors thanks to its four physical cores. However, Intel's quad-threaded Pentium Gold G5600 basically scored a tie with very similar performance. The G5600 even outperformed Intel's Core i3-7100, emphasizing the company's dramatic changes this generation.

The Pentium Gold family was no match for AMD's overclocked Ryzen 3 2200G, though. Perhaps surprisingly, the tuned Ryzen chip matched Intel's Core i3-8100, proving it has the multi-threaded chops to push our beefy graphics card in games that respond well to parallelization.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Civilization VI Graphics & AI, Dawn of War III

Civilization VI AI Test

Civilization VI's AI test measures CPU performance in a turn-based strategy game and tends to favor per-core performance, which is a mixture of IPC throughput and frequency.

Intel's Pentium models were bested by the stock Ryzen 3 2200G, though it's interesting that the Pentium Gold G5600 slid past last generation's Core i3-7100.

Civilization VI Graphics Test

The Core i3-8100 landed way out front. However, its $117 price tag is also way higher than any of the competing models.

At stock clock rates, the Pentium Gold G5600 led our more value-oriented test samples. Once we began overclocking, though, Ryzen 3 2200G jumped out ahead of the G5600.

As an aside, Ryzen 3 1300X features two quad-core CPU complexes tied together with AMD's Infinity Fabric. The company disables two cores in each CCX, yielding four active cores across the processor. In contrast, Ryzen 3 2200G employs one fully active CCX, so it doesn't suffer the same latency penalty that comes from communicating across the fabric. We know from past tests that this latency negatively affects gaming performance, so it was no surprise to find the 2200G beating the 1300X in certain titles at the same 3.9 GHz frequency.

Warhammer 40,000: Dawn of War III

Warhammer 40,000: DoW III's benchmark scales well with execution resources, though aggressive clock rates also provide a big benefit.

While the Pentium Gold G5600 delivered solid performance at its price point, Intel's G5400 only trailed by 2.5 FPS. That's not a bad trade-off for $30 in savings.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Far Cry 5, GTA: V & Hitman

Far Cry 5

Far Cry 5 placed the Coffee Lake-based Pentiums in front of their predecessors by a significant margin, which is impressive since both generations were launched at similar price points.

The Pentium Gold G5600 and G5400 beat AMD's stock Ryzen 3 processors. However, overclocking the Ryzens propelled them into the lead.

Grand Theft Auto V

Again, the Coffee Lake-based Pentiums beat their previous-gen equivalents.

This time around, though, AMD's Ryzen 3 2200G didn't need an overclock to establish dominance. Tuning improved its position even more.

Hitman

Hitman showed Intel's Pentium CPUs landing in a familiar order. Meanwhile, the stock Ryzen 3 2200G tumbled in our rankings. It's a good thing for AMD that overclocking is so effective.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Shadow Of War & Project CARS 2

Middle-earth: Shadow Of War

The $124 Core i3-8100 continued its dominance at the top of our chart. With that said, Middle-earth: Shadow of War doesn't scale particularly well, so there wasn't a ton of difference between our fastest and slowest CPUs.

Many of today's games are similarly constrained by graphics horsepower, which is important to remember when you're shopping for a budget-oriented chip. Just look at Ryzen 3 2200G's performance: overclocking didn't help it much at all.

Project CARS 2

Project CARS 2 is purportedly optimized for threading, and our test sequence did run better under Ryzen 3's four physical cores than the Pentium's two Hyper-Threaded cores. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Office & Productivity

Adobe Creative Cloud

Adobe's Creative Cloud suite is weighted towards single-threaded performance, a strength of Intel's architectures. Nevertheless, Ryzen 3 2200G finished ahead of the Pentium Gold G5600 by a few points. Some of AMD's advantage in these tests come as a result of the Spectre and Meltdown patches that have a big impact on Intel's performance in the Adobe test suite.

Pentium Gold G5600 still carved out a sizeable lead in a few lightly-threaded applications, such as Illustrator and InDesign. But AMD's Ryzen processors excelled in the Photoshop and After Effects workloads. 

Web Browser

The Coffee Lake-based Pentiums performed better in our web browser tests, which respond most readily to per-core performance.

While AMD's processors weren't as impressive at their stock settings, overclocking changed the story dramatically in Krakken and WebXPRT. Intel dominated the MotionMark tests that emphasize graphics performance (rather than JavaScript).

Productivity

The application start-up metric measures load-time snappiness in word processors, GIMP, and Web browsers, under warm- and cold-start conditions. The Coffee Lake-based Pentiums took an uncontested lead and easily dispatched the previous-gen Core i3-7100. 

Video conferencing measures performance in single- and multi-user applications that utilize the Windows Media Foundation for video playback and encoding. It also performs facial detection during the workload to model real-world usage. This task responds well to extra threads, so AMD's Ryzen 3 processors had no trouble establishing a lead over the Pentiums.

The photo editing benchmark measures performance with Futuremark's binaries that use the ImageMagick library. Common photo processing workloads also tend to be parallelized, so there was a big divide between the Pentium and Ryzen processors. A lack of support for AVX instructions compounded the Pentium's situation.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Rendering, Encoding & Compression

Rendering

Although Hyper-Threading is a great addition to Intel's Pentium family, the dual-core architecture cannot compete with true quad-core designs in heavily parallelized benchmarks like Cinebench and the Corona rendering test. Axing AVX support also hurt the Pentiums in Blender, which is optimized for CPUs able to accelerate AVX2 instructions.

Of course, Intel still enjoyed an advantage over AMD in our single-threaded rendering tests due to superior per-core performance. But did you notice that the Kaby Lake-based Core i3-7100 beat the newer Core i3-8100 in many of those lightly-threaded tasks? Although Intel doubled the number of physical cores on its Coffee Lake-based Core i3s, the company had to cut Core i3-8100's clock rate by 300 MHz in the process, slowing it down in certain benchmarks.

Encoding & Compression

Our threaded compression and decompression tests pull data directly from system memory, removing storage from the equation. This force-feeds the cores with information, giving quad-core models an advantage.

The difference between AMD's AVX-capable Ryzen processors and Intel's Pentiums was clear during our HandBrake and y-cruncher tests. AVX instructions are becoming more common in productivity applications, which makes Ryzen 3 2200G a better value in those types of workloads.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Final Analysis

Intel has a hard time adding features and functionality to its entry-level desktop CPUs without upsetting the balance of a carefully cultivated product stack. Improvements made to entry-level Pentium processors must extend upward through the Core i3, i5, and i7 families. After all, nobody is going to spend extra money on a higher-end chip unless it offers an appreciable advantage. 

The Kaby Lake-era Pentiums gained Hyper-Threading support, allowing two execution cores to operate on four threads simultaneously. Intel then followed up with new Core models with additional cores, maintaining the status quo. It's only a bummer that today's Pentium processors still lack other features, such as AVX support and unlocked ratio multipliers. While those differentiators make sense in the context of Intel's portfolio, they're liabilities compared to AMD's Ryzen 3 2200G.

Intel's Pentium family was largely unharmed by first-gen Ryzen CPUs because they landed at higher price points and lacked integrated graphics. But that Ryzen 3 2200G is a different beast. It sells for less than $100, similar to the Pentium Gold G5600, and wields a Radeon Vega graphics engine that decimates Intel's UHD Graphics 630. Add in AVX instruction support and an unlocked multiplier for true enthusiast appeal.

In the chart below, we plot gaming performance with both average frame rates and a geometric mean of the 99^th percentile frame times (a good indicator of smoothness), which we then convert into a frame-per-second measurement. We didn't include platform costs because these processors drop into value-minded platforms and come with decent stock coolers. It's noteworthy that we could overclock the Ryzen 3 2200G's CPU cores with AMD's stock thermal solution. However, you should buy something better if you also plan on tuning the chip's integrated Radeon Vega graphics. Also, bear in mind that we tested with an Nvidia GeForce GTX 1080 in order to alleviate graphics-imposed bottlenecks. Differences between our test subjects would shrink with more mainstream graphics cards installed.

In games, a $100 Ryzen 3 2200G trailed Intel's Pentium Gold G5600 just barely at its stock settings. But overclocking the Ryzen made it competitive with the $117 Core i3-8100. As such, we're doubling down on our recommendation to pair the Ryzen 3 2200G up with an add-in graphics card for gaming. And it's even more convincing across our application tests. The 2200G's four physical cores and AVX support provide superior performance in threaded workloads. The chip even fares well in many lightly-threaded tasks—and that's before we take overclocking into account.

Due to a slightly lower clock rate, the Pentium Gold G5400 can't quite match the G5600's performance. It doesn't trail by much in our gaming and application tests, though. Although the Pentium sold at a premium immediately after launch, it's now available for $64. That $30 savings is worth considering, particularly if you reinvest those funds into a faster graphics card or larger SSD. Intel's Pentium Gold G5400 looks like a great choice for builders on tight budgets, and AMD has nothing to compete against its price point. At least for now, the Pentium Gold G5400 reigns uncontested.

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPUs Content

Intel Core i7-8086K 40th Anniversary Model

Intel's 8086, the company's first processor to use its ubiquitous x86 instruction set architecture, debuted on June 8, 1978. Forty years later and by some stroke of fortuitous timing, Intel's desktop CPU portfolio is loaded with eighth-generation Core processors. So it was only fitting, then, that after a bit of prodding by a well-known chip analyst, Intel announced that it'd pay homage to the 8086 with a 40th-anniversary limited-edition Core i7-8086K.

Core i7-8086K is based on the same Coffee Lake architecture as Core i7-8700K, right down to its six Hyper-Threaded cores able to work on 12 threads concurrently. But it features a higher base frequency and more aggressive Turbo Boost bins, which tell us that Intel carefully picked out the best dies to use in these chips. This is the first Intel processor to ship with a 5 GHz Turbo Boost bin, matching AMD's record with the FX-9590. And if you're only looking at clock rate, the -8086K represents a 1000x multiplication of the original 8086's 5 MHz frequency.

Incidentally, the -8086K is also Intel's first six-core processor with a 4 GHz base frequency, though that specification isn't as eye-catching.

Intel kicked off its anniversary celebration with a giveaway of 8086 Core i7-8086Ks. If you didn't win one, you'll have to purchase the processor like we did. Your window of opportunity won't be large, though: our sources confirm a production run of just 50,000 units. We expect collector's items to sport premium pricing, and Intel doesn't disappoint in that department. As of this writing, the -8086K sells for $75 more than the once-flagship Core i7-8700K.

So what is this processor's appeal, other than the obvious nostalgia? Core i7-8086K comes from a higher-quality bin than Core i7-8700K, so enthusiasts with deep pockets can expect to receive the very best example of Coffee Lake silicon available. Of course, most folks won't consider the extra $75 worth paying for moderate gains at stock clock rates. But again, this is a limited-edition piece of hardware steeped in history.

Intel Core i7-8600K

The 6C/12T Core i7-8086K is manufactured on Intel's 14nm++ process, just like its other Coffee Lake CPUs. Like the company's Core i7-8700K, its 95W Core i7-8086K also features 13MB of L3 cache, support for up to 64GB of dual-channel memory at DDR4-2666, an unlocked multiplier to facilitate overclocking, and Intel's integrated UHD Graphics 630 engine that can boost up to 1.2 GHz. For more information about the Coffee Lake architecture, check out our Core i7-8700K review.