Apple's fans know all about waiting. The company tends to polish features, especially in hardware, that some other laptop vendors release faster. The MacBook Pro with PROLED XDR Display ends the wait for two long-requested elements: OLED and touch, both on the platform for the very first time.

That comes at a price. The 16-inch version we have on review starts at $3,399 and is $7,899 as configured. 14-inch systems start at $2,899. Those are $400 bumps over the existing MacBook Pros with identical chips, making the OLED panel a pricey upgrade.

But those who can afford these high-end systems will get the best display on a Mac, if not laptops in general. Add in a significant reduction in weight and improved speakers, and you're getting a lot of small tweaks that add up over what was already one of the best laptops, even if many of the aesthetics haven't changed.

Design

From a distance, this new MacBook Pro doesn't look all that different from the existing models, which are still on the market, which may upset some people looking for a more sweeping redesign. They come in the same colors (space black, like our review unit, and silver) and they largely have the same design. Technically, the new OLED display stack allows for a thinner system, but our 16-inch review unit, at 16 mm thick, is just 0.8 mm shorter than an equivalent 16-inch MacBook Pro with Liquid Retina screen.

The aesthetic differences show up once you open the laptop. For one, the notch in the display is gone with the PROLED XDR models, replaced with the Dynamic Island from the latest iPhones. Let us rejoice; while the notch never bothered me on phones, it always felt in the way, dipping into the macOS menu bar. This isn't perfect, but it feels far more natural and has added functionality (see "Dynamic Island and macOS Touch Features" below).

Then there's the glowing Apple logo on the back of the lid. We haven't seen this for years, and while I could take or leave it, it's definitely going to make some fans nostalgic. The old glowing Mac logos used light from LCD backlights, getting the effect with no extra power. Because OLED panels only light up pixels when they're needed, Apple is making an aesthetic choice to power the logo with a bit of extra electricity this time. The glow matches the brightness of the display, which is a fun touch. I didn't see anything in settings to disable it.

Apple's bezels are thinner, which has let Apple stretch the display to 16.3 inches on this new 16-inch model (it's still 14.2 inches on the smaller one). The displays are now rounded on all four sides. On the Liquid Retina displays, they're rounded on top to match the curves in the chassis, but are squared on the bottom. It's not a huge change, but it does cut some pixels on the bottom of the screen and looks a bit like an iPad.

Otherwise, the system looks largely the same. You have the same ports: MagSafe 3, two Thunderbolt 5 ports, and a 3.5 mm headphone jack on the left; and an SD card reader, an additional Thunderbolt port, and HDMI output on the right.

But the OLED laptop feels different when you pick it up or put it in a backpack. Changes to cooling in the new 16-inch MacBook Pro with PROLED XDR and the switch to an OLED screen get the laptop down to 4.1 pounds with M5 Max (4.0 pounds with M5 Pro), whereas the Liquid Retina versions are 4.7 pounds. You can really feel the half-pound difference. (The 14-incher is down from 3.5 pounds to 3 pounds, too.)

And that's lighter than some of its new competition, too. Microsoft's Surface Laptop Ultra with Nvidia's RTX Spark chip is 4.4 pounds and thicker at 0.75 inches. The Lenovo Pro 9n, also with RTX Spark and its own OLED screen, is a lighter 3.64 pounds and is thicker at 16.7 mm.

There has been talk of a sturdier hinge, but it does still wobble when you touch it. No one has gotten around that issue. Comparing it to my personal MacBook Pro with the old design, the display goes back ever so slightly more, which is nice for touch but not a game-changer. It would be nice to see a 180-degree hinge for people who really want to write or draw on a flat surface (Apple Pencil support would be nice for that, too). For now, people who want that will need to stick to an iPad.

I spent some of the time reviewing this laptop thinking of Microsoft's Surface Laptop Studio, a touch-focused workstation with a hinge that lets the screen sit in three positions: a standard laptop, pulled forward over the keyboard, or lying like a tablet. That didn't succeed (neither did similar designs from Acer), but it took a stab at changing how a touch-based laptop worked. The new MacBook Pro design isn't a radical rethink of the ergonomics that have plagued touch laptops for over a decade. It's just a powerful Mac that happens to have a touchscreen.

Specifications

PROLED XDR Display

We had to wait a long time to get OLED on a Mac. PROLED XDR is a silly name, but I'll give Apple this: The screen is absolutely stunning. It's the best screen I've ever seen on a notebook, and it makes me jealous when I look back at my personal system with plain old mini LED. Everything from the default wallpaper in macOS Golden Gate to videos and photos pop on this display. The blacks are deep, the colors pop, and it can get intensely bright.

Our review unit came with the 16.3-inch (slightly enlarged thanks to thinner bezels from 16.2-inches on non-PROLED models) with 3472-by-2266 native resolution, a variable refresh rate of 10 - 120 Hz, and, of course, touch support.

I used the PROLED XDR display to watch the finale of the latest season of Slow Horses. In one scene, where the mastermind of the season's plot shows up to a restaurant to meet with the head of MI5, I was wowed by the deep blacks. The restaurant was sparsely lit, and both characters were wearing black coats. The display brought out the highlights in the environment, as well as a few red accent walls.

In our instrumented testing, the PROLED XDR screen covered 83.6% of the DCI-P3 color gamut by volume and 118% sRGB, actually slightly lower than the mini LED version. I've seen them side-by-side. Trust my eyes; there's no contest here.

The MacBook Pro with PROLED XDR measured 542.6 nits in our SDR brightness test, falling below the Surface Laptop Ultra, with its near-blinding 944 nits with its mini LED display. But again, PROLED simply looks better. The Surface Laptop Ultra has some light bleed and not enough local dimming zones to catch up.

The OLED panel on the Lenovo Pro 9n is the most vivid on each gamut by volume. I think it's a bit oversaturated, but if you prefer extreme pops of color, you may prefer that screen. The tandem OLED on the Dell XPS 16 was also more vivid on measurements, but was far less bright, which mutes the best aspects of OLED.

The PROLED screen showed how bright it could get with HDR content, outputting 1,964 nits in our 40% HDR test.

While I love how the display looks out of the box, Apple is offering more reference modes than ever before on a laptop, including various video, cinema, web and photography standards, and even DICOM medical imaging at 350 and 600 nits. There's also a display calibrator to fine-tune white points or perform full recalibration with external measurement tools.

Apple is far from the first company to add OLED (or, in this case, tandem OLED) to a laptop, but this stands out among the best OLED laptop displays I've seen in recent memory. The price hike to get it — $400 over mini LED models with the same chips, will be out of reach for many (though that hike also shaves half a pound off the weight). I desperately hope Apple starts bringing its brand of OLED down to the rest of the MacBook Pro and MacBook Air line, too. After all, the iPad Pro and iPad Mini both have it.

We tested the MacBook Pro with a standard screen. While I haven't spent significant time with a PROLED with the matte nano-texture display, I can tell you that this looks so good with a glossy finish that I'm not sure I could risk letting the matte displays (which I usually love!) take away any of the color saturation or context on display here. And as far as reflections go, they were minimal, even next to a window.

Dynamic Island and macOS Touch Features

Outside of OLED, there are two other additions to the MacBook Pro's screen.

The first is the Dynamic Island, a group of black pixels around the punch-hole webcam that can be used to show system alerts, notifications, and ongoing activities. (If you plug in a monitor, Dynamic Island activities show up to the left of your menu bar apps and icons.) It matches modern iPhones, and it's a huge change over the notch in other Macs, which took up far too much space in the menu bar. Now, the area around the camera is at least useful, though the camera still doesn't offer Face ID.

One of my biggest annoyances with the old design was that your mouse could fall behind the notch, with no indicator of where it was. The mouse can still move behind the Dynamic Island, but now the black area grows to indicate where the pointer is, and even wiggles if you make very small movements, to remind you where the cursor is.

The other much-discussed addition is the touchscreen. Apple hasn't reinvented the wheel here. Like most Windows and Chromebook makers, it added touch functionality onto a standard laptop. Apple's announcement mentions that the "touch sensor is built directly into the display panel — not layered on top of it," suggesting the use of technology similar to the iPhone and flagship Windows laptops.

Apple has made slight adjustments to macOS 27 Golden Gate to accommodate touch. My personal favorite is that the "traffic lights" for maximizing, minimizing, or closing a window expand when you tap on them, to ensure you can hit the right one. When you click on a menu with your finger, the spacing is larger for easier touch targets. Like the iPhone, you can swipe down from the dynamic Island to invoke Siri. You can also flick windows to the sides to organize your windows, though you can already do that with a mouse or keyboard.

Many gestures that work on the trackpad also work on the touchscreen, like pinch to zoom, accessing Mission Control (swiping up with three fingers), showing Notification Center (swiping left from the edge of the screen), and moving between virtual desktops by swiping left or right with three fingers. Many apps have already been optimized for touch, but if they haven't, touch translates to a click, so it all still works fairly well.

One of my many experiments with the touch screen included swiping across the timeline in Final Cut Pro. But in a sample project I put together, I didn't find it any better than using the keyboard and mouse — just another option. In Maxon's Cinema 4D, I felt the same way moving a model car and adjusting virtual cameras. It's an option, but it's no better than the existing way of doing things.

Perhaps my favorite use was in Blender, making adjustments to a 3D model I was sculpting. You could do everything I did with a mouse, but some features, like changing the angle of the sculpture, felt very natural with touch. That being said, after a few minutes, my arm was tired. I couldn't see using this all the time, as there are no changes to ergonomics over the usual laptop design.

For those who never want to use touch, you can turn it off in disability settings. That being said, as someone who would likely only use it for occasional scrolls, I don't see the harm in having it on, even if I don't see myself ever pulling to refresh on Safari when CMD + R is right there. (I do wish Apple included the polishing cloth with the glossy screen on our review unit and not just matte screens, however. Touch screens get fingerprints!).

I found touch most useful for iPad apps on the App Store that work natively on Mac. For instance, my go-to podcast app is Overcast, which is great on my iPhone but designed for touch on Mac. While I'm very used to navigating it with a mouse, the choice to use touch made sense here. It's also great for the iPhone simulator in Xcode, where developers can try out touch-based apps on their touchscreen Mac instead of using a pointer. The same goes for iPhone Mirroring.

There's no touch tutorial. Touch is clearly a feature on these new MacBook Pros, but not the feature. Keyboard and mouse are clearly still king.

Audio

The six-speaker sound system in the 16-inch MacBook Pro's new design, complete with updated woofers and tweeters, keeps the crown in Apple’s camp for the best laptop speakers I've ever heard.

When I listened to Justice's "Flower in a Hurricane" with the MacBook Pro, I was floored by the drums. While the rest of the speakers don't sound superbly better than the existing Pros, the low-end has improved, and deep drums and bass easily filled my apartment with deep, well-rounded sound. Everything sounded incredible, from the audio to the synths. This may be the best way I've heard this song since I've had it on a loop after its debut in a Grand Theft Auto VI trailer.

Apple has made some adjustments to how the speaker sounds with the lid shut, too. It's surprisingly good, with The Beaches' cover of "I Ran (So Far Away)" getting surprisingly loud, with the laptop plugged into a Thunderbolt dock and the lid shut, though the drums in that song weren't as snappy. Everything was slightly dampened until I opened the lid, but honestly, I could use it like this. It's still better than many other laptop speakers, with better bass.

Productivity Performance

We tested the new MacBook Pro with an M5 Max, 128GB of unified memory, and a 2TB SSD. That's the same chip and RAM as in the 14-inch MacBook Pro we tested earlier this year, though that had a 4TB drive.

The main competitors we looked at featured Nvidia's RTX Spark N1X chips, including the Surface Laptop Ultra (20-core version) and the Lenovo Pro 9n (18-core version). We also tossed in the Dell XPS 16 with a Core X7 358H — another 16-inch system with a Tandem OLED.

Unsurprisingly, this MacBook Pro can tear through most people's everyday workloads. Apple has re-tooled the cooling in this system, and it was often close to, if not ahead of, the M5 Max we tested earlier this year in a 14-inch chassis.

On Geekbench 6, the 16-inch MacBook Pro with PROLED XDR earned a score of 4,353 and a multi-core score of 29,430, coming close to the other M5 Max machine. Both of those scores are far ahead of the RTX Spark laptops and the Intel-based Dell.

The 2TB SSD copied roughly 25GB of files at a rate of 1,917.69 MBps, which is fast, though the 4TB SSD in the MacBook Pro we tested earlier this year was far faster. The Surface Laptop Ultra was ever-so-slightly ahead at 1,998.59 MBps.

On Handbrake, the 16-inch MacBook Pro transcoded a 4K video to 1080p in 1 minute and 43 seconds, the fastest of the bunch.

On the Xcode benchmark, in which Macs compile large code bases, the new 16-inch MacBook Pro beat the older 14-inch model with the same M5 Max by 7 seconds, completing the task in 80 seconds.

Our Cinebench 2026 stress test has systems go through ten runs of multi-core rendering, which is particularly stressful. In the automatic performance mode, Apple's default out of the box on macOS 27, the system started at a score of 9,080, before eventually settling in the 8,050's due to throttling. With high performance mode, which runs the fans harder and louder, the Mac started at a score of 9,647 and was far more stable, with its lowest score (9,479) at the very end. If you need to get some tough work done faster, switching to high performance mode could help, as long as you're OK with fan noise.

Gaming and Graphics

The MacBook Pro with PROLED XDR display came with the same 40-core M5 Max GPU as the MacBook Pro we reviewed earlier this year.

In the new chassis, the M5 Max scored 4,355 on 3DMark Steel Nomad, beating the old system (3,865) as well as both configurations of the RTX Spark.

Running Cyberpunk 2077's benchmark at 1920 x 1200 using our typical test settings (Ray tracing ultra), the game ran at 36.24 frames per second. This is slightly faster than the 14-inch we tested earlier this year with the existing chassis, at 30.84 FPS. With the "For This Mac" preset, including MetalFX and a resolution of 2316 x 1448, the benchmark ran at 94.20 FPS.

Keyboard and Touchpad

Of all the things that haven't changed on the MacBook Pro, the most familiar are the keyboard and touchpad. That may be of some comfort to stalwarts of indirect input methods.

I was blazing fast on the keyboard, hitting 131 words per minute with a 99% accuracy rate on the monkeytype.com typing test. Both stats are on the high side for me, but I also can't blame Apple for making a comfortable keyboard with chiclet keys and a smart layout. I will praise any company that uses an inverted T for arrow keys as long as I live. Apple may not offer the deepest travel of the laptop market, but it does the job.

The massive trackpad features smooth glass, comfortable haptics, and intuitive gesture support in macOS. The 16-inch MacBook Pro's touchpad is luxuriously huge. After years of Windows laptops adding haptic trackpads to catch up, Apple's touchpad still feels great.

Upgradeability

While there are just four Torx Screws on the bottom of the MacBook Pro, there isn't much in the way of upgrading. Apple puts the RAM in the chip, and the storage is soldered down. If you need a repair, your best bet is probably Apple's Genius Bar.

Like other Macs, you should make sure you get the spec you need when you buy. Though, as you'll see in the configuration section, that could cost you quite a bit for some RAM and storage upgrades. Unfortunately, that's not unusual given the price of memory in the greater industry right now.

Battery Life

Apple's 99.6 WHr battery allows for a longer-lasting OLED laptop than its competitors, but in our testing, mini-LED model still lasts longer. On our battery test, laptops browse the web, run OpenGL tests, and stream video with the display set to 150 nits of brightness.

The 16-inch MacBook Pro with the PROLED XDR display lasted 15 hours and 57 minutes, far ahead of the OLED screen on the Lenovo Pro 9n (10:30) and the tandem OLED panel on the Dell XPS 16 (13:03). It also beat the mini-LED screen on the Surface Laptop Ultra.

The nearly 16-hour runtime is slightly short of Apple's "up to 17 hours" claim for web browsing. The only laptop that did better on our battery test was the 14-inch M5 Max laptop we tested earlier this year. That machine, with a smaller, mini-LED screen, ran for just shy of 18 hours.

Heat

The MacBook Pro can get a bit toasty under heavy load, though it's nothing I would deem unusable. Heavy workloads are probably best on a desk, though. New cooling changes the direction of the airflow in the machine, moving it out and through the back rather than out from vents near the display, in order to extend the display's lifespan.

During our Cinebench 2026 stress test, the keyboard measured 106.6 degrees Fahrenheit, which is definitely hot enough to notice. The touchpad was a cooler 88 F, while the hottest point on the bottom measured 107 F.

Our usual monitoring was not able to identify the heat on the CPU yet (this may require some software updates), but the overall temperature from the computer's other sensors was an average of 51 degrees Celsius.

Webcam

The 12-megapixel Center Stage webcam on the MacBook Pro will serve you well even on the most professional call. It looked great in meetings, even in one difficult lighting scenario when I had a lamp on one side and post-Daylight Savings Time early evening light coming in from a window next to me. The system gracefully balanced the mix of warm yellow and the products of blue hour.

I could easily smooth that mix out using the MacBook Pro camera's studio light or edge light effects, too.

Software and Warranty

We tested this new MacBook Pro with macOS 27 Golden Gate, which we highlighted on the Mac Studio, with new Apple Intelligence features, and the new Siri based on Apple Foundation Models and Gemini.

Apple's software suite includes Safari, Messages, Photos, Maps, Notes, and Reminders. Its services, including News and Apple TV, also have applications. Pages, Keynote, and Numbers are still in the dock by default, but they're the Creator Studio subscription versions, which require a subscription to get full features, including AI functionality, royalty-free graphics, and additional templates.

Apple sells the MacBook Pro with PROLED XDR Display with a one-year warranty. AppleCare+ extends protection for $17.99 per month or $179.99 per year, and also adds accidental damage and drop coverage.

Configurations

We reviewed the MacBook Pro with PROLED XDR Display in a high-end, $7,399 configuration with an M5 Max (18 CPU cores, 40 GPU cores), 128GB of memory, and 2TB of storage.

The cheapest versions of these machines are a $2,899 14-inch system with a cut-down M5 Pro featuring 15 CPU cores and 16 GPU Cores or a $3,399 16-inch laptop with the M5 Pro featuring an 18-core CPU and 20-core GPU. Both base configurations come with 24GB of RAM and 1TB of storage, as well as the glossy screen option. (Adding the matte "Nano-texture" glass adds $150 and also nets you a polishing cloth, which isn't included with the glossy screen.)

Moving to the M5 Max moves you to $4,499 on the 14-inch laptop and $4,799 on the 16-incher for an 18-core CPU and a cut-down 32-core GPU. The full 40-core GPU is another $300.

RAM starts at 24GB, and costs $600 to goto 48GB or, on the 16-incher, $1,000 for 64GB. There's also 36GB and 128GB RAM options that also require 2TB of storage and an M5 Max (like our review unit). Storage prices are also astronomical, with a $1,000 jump between the 2TB and 4TB tier. (If you want 16TB, which I think will be a relatively rare choice, it's a $7,000 bump that requires the high-end M5 Max).

These are exceptionally expensive computers. Apple's 14- and 16-inch MacBook Pros without the OLED screens are still there, however, starting at $1,999 for a 14-inch notebook with M6.

Apple sells the systems at full price, or divided up into 12 monthly payments through an installment plan. They're also available for lease through Apple Upgrade.

Bottom Line

The MacBook Pro with PROLED XDR display isn't a radical redesign, but it has a bunch of small changes that add up to significant improvements over previous models. The tandem OLED screen is a treat to look at, and the weight reduction is significant for anyone carrying their laptop around on the go. Even the deeper bass from the speakers is a real treat.

Performance from the M5 Max is still strong, though if you're doing simple work, you can get much better single-core performance out of an M6 Macbook Pro (though you'll give up everything else).

With no major redesign, however, Apple hasn't done anything to make touching the laptop any more comfortable than its competitors. It ultimately does get tiring. That being said, the few gestures and adjustments Apple has made to macOS are helpful, if not revolutionary. Touch really is here because it's table stakes at this point. Some will use it and others won't. It's not the selling point, other than that it's a debut for the feature.

Far more Mac user deserves a screen this good, and I hope that some flavor of OLED, even if it's not tandem OLED, eventually makes it to the entry-level MacBook Pro and even the MacBook Air. Touch, too, could be popular with younger people who grew up with all touch screens and go for the Air or the MacBook Neo.

But the starting price for the systems with PROLED displays are high, and upgrades make them exceptionally expensive. For now, the best screens in the Mac lineup (and arguably the industry) are limited to just these systems. We've waited a long time for OLED on the Mac, and for many, they'll still be waiting until it gets more affordable.

Those with the deep pockets to make the investment, though, are in for one heck of a screen.

It’s that time of the year again, which means that Apple has released yet another refreshed 14-inch MacBook Pro using a legacy chassis. While it looks the same as last year’s M5 models on the outside, it has some key hardware upgrades underneath that address performance and connectivity.

This time around, we have a MacBook Pro using the new M6 chip that was first introduced in the refreshed Mac Mini. In addition, Apple has outfitted the MacBook Pro with Wi-Fi 7 and Bluetooth 6 connectivity courtesy of the in-house Apple N1 chip.

Thanks to industry-wide price hikes on memory and storage, Apple’s 14-inch MacBook Pro now starts at $1,999.

Design of the MacBook Pro (M6, 14-inch)

Apple has been using the same design for its entry-level MacBook Pro since 2021, and there are no changes this year (if you want a slightly refreshed design, check out the touch screen-enabled 14- and 16-inch MacBook Pros). It’s still a good-looking laptop, but it’s showing its age. Our MacBook Pro review unit came finished in silver instead of space black. While space black looks great, it’s a fingerprint magnet, so many people might prefer the lighter finish, which better hides your skin oils.

The notch at the top of the display now seems archaic, especially after seeing the Dynamic Island on the pricier MacBook Pros. The notch is always staring back at you, even though I have mostly gotten used to it over the years. The main thing going in the notch’s favor is that it resides in the same space as the menu bar, which limits the visual annoyance.

My only other niggle with the design is one that I’ve harped about in previous MacBook Pro reviews: the cutout in the keyboard deck that aids in lifting the lid has sharp corners, which can jab your palms while typing.

As with previous MacBook Pro generations, you’ll find three Thunderbolt 4 ports (Thunderbolt 5 is not available on M6 models), a full-size SDXC card reader, HDMI 2.1, a 3.5 mm audio jack, and a MagSafe 3 connector.

Our 14-inch MacBook review unit measures 12.31 x 8.71 x 0.61 inches and weighs 3.4 pounds, the same as last year’s model with the M5. For comparison, the Gigabyte Aero X16 measures 14 x 9.9 x 0.78 inches and 4.2 pounds, while the Dell XPS 14 comes in at 12.19 x 8.26 x 0.6 inches and weighs 3 pounds.

MacBook Pro Specifications (M6, 14-inch)

Productivity and Graphics Performance on the MacBook Pro (M5, 14-inch)

Our 14-inch MacBook Pro review unit features a 12-core CPU and 12-core GPU version of Apple’s new M6 SoC. This is the same configuration as the Mac Mini that we reviewed in September, except that our laptop review unit includes 16GB of unified LPDDR5x memory compared to 24GB. Rounding out the specs is a standard 1TB SSD.

Slow and steady wins the race in silicon advances. In every productivity benchmark, our M6 MacBook Pro not only outshone the previous-generation M5 MacBook Pro, but it also trounced the prosumer Gigabyte Aero X16 (AMD Ryzen AI 7 350) and Dell XPS 14 (Intel Core Ultra X7 358H) Windows 11 laptops that we pitted it against.

Starting out with the Geekbench 6 synthetic CPU benchmark, our M6 review unit put up a modest 10% gain in single-core performance over the M5 (4,752 versus 4,288). Switching to multi-core performance, that performance delta grew to just over 18%. Of the two Windows 11 laptops, the XPS 14 came the closest, putting up single- and multi-core scores of 2,867 and 16,927, respectively.

Apple appears to have held the line when it comes to SSD performance year-over-year. 2025's M5 MacBook Pro managed 1917.35 MBps in our in-house 25GB file transfer test. Things changed little this year, as our M6 MacBook Pro came in at 1914.23 MBps. The Aero X16 carried the baton for the Windows 11 field with a result of 1,480.57 MBps.

Our Handbrake test consists of transcoding a 4K video file to 1080p. The new M6 MacBook Pro review finished the job in a brisk 2 minutes and 50 seconds, while the M5 version took 3 minutes and 31 seconds to complete the task.

Our stress test for productivity laptops centers around Cinebench 2026, which we run 10 times back-to-back, tallying multi-core scores along the way. The M6 MacBook Pro started with a score of 5,026.55 and hovered around 5,440 for the remainder of the runs.

The M5 includes a 12-core GPU (each core features a dedicated neural accelerator), which promises up to a 30 percent uplift in performance compared to the M5.

Although the base MacBook Pro isn’t what we’d consider a gaming beast, it did manage 53.7 frames per second (FPS) at 1800 x 1125 resolution in Cyberpunk 2077 using the “For this Mac” preset and 32.66 FPS at 3024 x 1890. This compares to 47.65 FPS and 31.47 FPS, respectively, for the M5 MacBook Pro.

Display on the MacBook Pro (M6, 14-inch)

Apple’s M6-equipped 14-inch MacBook Pro continues to use an IPS display with mini LED backlighting. As before, you get a 3,024 x 1,964 resolution panel with a 120 Hz refresh rate.

While the mini LED panel looks good, it’s not nearly as capable as the touch-enabled PROLED XDR display found on its pricier siblings. Although you can equip the MacBook Pro with a nano-texture display for $150, our review unit didn’t have it.

Being that I’m a huge fan of Sylvester Stallone’s Rocky films (well, the first four), I pulled up the second trailer for I Play Rocky, which is set to release later this month. The design aesthetic of the film is, of course, distinctly 1970s, and the mini LED panel did a good job of presenting the muted colors (mostly warm earth tones).

As I’m still in shock over the passing of Patrick Stewart, I watched a sizable portion of Star Trek: First Contact on the screen. Again, the display performed well, with major kudos during the opening battle scene where Federation ships are battling the Borg. The orange explosions, red phaser blasts, and the Borg’s green cutting beams vividly danced across the screen. The blackness of space wasn’t quite as deep as I witnessed with an OLED screen side-by-side, but it’s still good.

Our MacBook Pro review unit was even brighter than last year’s model, if only by a slim margin (571.4 nits versus 558 nits). That performance was still well ahead of the OLED panels in the Windows laptops.

Those similarities extended to color performance, which tracked nearly identically to the preceding M5 MacBook Pro, which was a dead ringer for its M4 predecessor. Our instrumented tests showed 81.5 percent coverage of DCI-P3 and 115.1 percent of the sRGB. Although the XPS 14’s OLED panel was no match for brightness (358 nits), it pipped the MacBook Pro in color performance (89.7 percent DCI-P3, 126.6 percent sRGB).

Keyboard and Touchpad on the Apple MacBook Pro (M6, 14-inch)

As with all the other exterior design elements of the M6 MacBook Pro, the keyboard remains unchanged. To me, that’s a good thing, as it’s one of the things I love the most about Apple laptops. As someone who lived through Apple’s horrible butterfly keys on the 12-inch MacBook and subsequent MacBook Pro models, the return to the venerable switches was a return to sanity on Apple’s part.

Apple isn’t the type of company that would equip its laptops with RGB keyboards, so you’ll get single-zone white LED key backlighting. Although the thought of a MacBook Pro with RGB lighting brings a slight smirk to my face, I’m perfectly fine with Apple’s monochromatic approach.

I used the keyhero.com typing test, as I do with all laptops I review, and settled in with 93 words per minute at 95 percent accuracy. Those figures track with what I achieve with other Apple and Windows laptops.

Like all MacBooks (except for the entry-level Neo), our MacBook Pro features a large haptic trackpad that measures 5 x 3 inches. The glass surface makes gliding with your finger or thumb easy-peasy, and the simulated “click” is satisfying.

Audio on the Apple MacBook Pro (M6, 14-inch)

We’re sure you’re noticing a theme in this review, but the audio system in our MacBook Pro review unit remains unchanged compared to its earlier M5 and M4 counterparts. What that means is that you’ll find a six-speaker system that includes support for major audio formats like Dolby Digital, Dolby Atmos, Dolby Digital Plus, and Apple Spatial Audio.

I’m a sucker for acoustic guitar, and one of the greats of the 2000s is John Mayer. Perhaps my favorite John Mayer song is “3x5” from his debut album, which my girlfriend at the time (now wife) and I played over and over again and witnessed several times in concert. On the MacBook Pro’s speakers, the acoustic guitar packs just as much of a punch as John’s breathy vocals pelting out vivid lyrics like, "Today, skies are painted colors of a cowboy cliché."

I got so lost in the audio experience that I found myself listening to the entire Room for Squares album and being transported back to my college days.

Upgradeability of the Apple MacBook Pro (M6, 14-inch)

While most Windows laptops give you some semblance of upgradeability, allowing you to upgrade at the very least storage, Wi-Fi/Bluetooth, and sometimes memory, you’ll find no such luxuries in MacBook land. There are four pentalobe screws on the bottom of the laptop, but once you remove the cover, there’s nothing of note that’s upgradeable. The M6 chip with its integrated 16GB of memory can’t be replaced, nor can the 1TB SSD – both packages are soldered onto the motherboard. As with all MacBooks, you’ll need to settle on how much RAM and storage you want at purchase time. If you get it wrong, you’ll have to live with that config until it’s time to upgrade your laptop again.

Battery Life on the Apple MacBook Pro (M6, 14-inch)

If there’s one thing that you can count on when it comes to MacBooks, it’s battery life. It’s pretty common for MacBooks to last for 18 or more hours in our battery life test, which consists of browsing the web, running OpenGL tests, and streaming video at 150 nits of brightness.

Our MacBook Pro review unit was no exception, lasting 19 hours and 5 minutes thanks to its 72 WHr battery. This was nearly a one-hour improvement over its M5 counterpart, which is a testament to the efficiency gains that Apple is able to achieve with the 2-nanometer process node.

Heat on the Apple MacBook Pro (M6, 14-inch)

We measured skin temperature on the MacBook Pro during our Cinebench 2026 stress test. Unlike the cheaper MacBook Airs, which are fanless, the MacBook Pro’s two internal fans kick into gear quite quickly to dissipate heat.

The hottest point was near the left exhaust vent, which measured 106 degrees Fahrenheit, with 105 F seen between the G and H keys. Given its distance from the toasty parts of the laptop, the touchpad remained cool to the touch at 78 F. The touchpad was much cooler at 82.8 F.

We use TG Pro to measure internal temperatures while running the Cinebench 2026 stress test. While TG Pro doesn’t have full support for the M6 yet, the Power Manager dies reported peak temperatures of 80 degrees Celsius (C) for the SoC. The overall system temperature, which takes an average of several components (i.e., power supply, battery, SSD, logic board, etc.), was 43 degrees C.

Webcam on the Apple MacBook Pro (M6, 14-inch)

The 12MP Center Stage camera that’s been featured in the last two MacBook Pro generations has made a reappearance this year. All of the superlatives that I had for the M5 MacBook Pro are relevant to my current review unit. Namely, image quality was impressive across the board, in both bright and low-light conditions.

Unlike some other laptop webcams that skew my darker skin color or apply smoothing algorithms that make my skin look like a Tootsie Roll, the camera picked up the wrinkles around my eyes and accurately represented the grays in my beard.

Software on the Apple MacBook Pro (M6, 14-inch)

The M6 MacBook Pro ships with macOS 27 Golden Gate preinstalled. You can liken the shift from macOS 26 Tahoe to macOS 27 Golden Gate to the OS X 10.5 Leopard to Mac OS X 10.6 Snow Leopard transition. In other words, macOS 27 is more about making the operating system more reliable and faster rather than bombarding users with new features. You can read more about macOS 27 Golden Gate in my M6 Mac Mini review.

Just like Windows 11, macOS comes with a slew of apps preinstalled upon first boot. Some of the standouts include GarageBand, Apple’s productivity suite (Keynote, Pages, Numbers), Apple TV, iMovie, and Home (Apple’s smart home application). With the exception of Apple TV, the other apps are productivity utilities meant to enhance the user experience and get work done. You won’t find typical bloatware like app/shortcuts for services like Dropbox or antivirus programs like McAfee as you would on Windows 11 laptops.

Apple MacBook Pro (M6, 14-inch) Configurations and Warranty

When opting for the 14-inch MacBook Pro with M6 processor, you’re looking at a starting price of $1,999. In addition to the M6 processor, you get 16GB of RAM and a 1TB SSD (the configuration of our review unit).

From there, you can opt for 24GB (+$200) or 32AM (+$400) of RAM, and 2TB (+$500) or 4TB (+$1,500) of storage. There are also two standalone options that you can add: the nano-texture display for $150 and a 96-watt power adapter for an extra $20.

The MacBook Pro has a standard one-year warranty. However, you can pick up AppleCare+ for $99.99/year or $9.99/month. The coverage adds unlimited repairs covering accidents, 24/7 priority tech support, and battery replacement.

Bottom Line

When it comes to everything you see on the outside, Apple’s latest 14-inch MacBook Pro is more of the same: same display, same chassis, same keyboard, and the same great build quality. But true to form, Apple has steadily focused on improving performance year-over-year, and we see that again with the new M6 SoC.

You’ll see anywhere from a 10% to 20% uplift across the board, depending on whether you’re dealing with single- or multi-threaded apps. On top of that, efficiency gains have netted nearly an additional hour of battery life over its M5-powered predecessor. And Apple has finally given the MacBook Pro Wi-Fi 7 and Bluetooth 6.0 support.

The M6 MacBook Pro doesn’t have the OLED touchscreen, more svelte design, or Dynamic Island of the all-new MacBook Pros, but it also started at a price that is $900 less than those systems. With that said, the $1,999 price of our review unit is still a lot of money to ask for a laptop featuring just 16GB of RAM and a 1TB SSD. And you don’t even get an OLED screen at that price.

But given the price of memory and storage these days, $1,999 isn’t exactly out of the ordinary. A Dell XPS 14 configured with an Intel Core Ultra 5 325, 16GB of RAM, 1TB SSD, and a 2.8K OLED display is $2,549. We are truly living in unprecedented times, and while it’s hard to reset our expectations concerning the price of tech, the M6 MacBook Pro remains a solid entry in a highly competitive market.

Apple is about to feel the squeeze from rising memory prices, with the cost of RAM climbing just as the company heads into its busiest shopping season. According to Reuters, Apple expects to pay more for memory in the December quarter, with the CEO, John Ternus, describing the current price increases as “unprecedented. The company has so far managed to soften the blow by using memory bought at lower prices, cutting costs elsewhere and raising prices on some products. However, it is reported that its manufacturing bill has climbed by $33 billion in just a few months.

But that cushion is getting thinner, and Apple expects memory purchases to become even more expensive into 2027. “What we're seeing with memory pricing right now is unprecedented. And as we said earlier in the September quarter, we did see higher memory costs as compared to the June quarter, which were partly offset by the benefit of carrying inventory and lower costs on certain non-memory components. Now for the December quarter, we expect even higher memory costs, partly offset by a few factors. So as you know, we changed the price on many of our products, including Mac and iPad at the end of June and iPhone in September, and these price increases provided some offset.” Ternus said during Apple’s November 2 earnings call.

Memory is used across Apple's hardware lineup, so sustained price increases could affect manufacturing cost for several product categories—although Apple has not announced another round of price increases. Absorbing higher components could, however, indefinitely become difficult. Dave Nicholson, Chief Research Officer at the Futurum Group, also identified memory prices as a growing pressure point for Apple's margins and hardware pricing through the holiday season and into 2027. According to him, Apple is in a different position from companies pouring hundreds of billions of dollars into AI data center infrastructure. The impact of higher component costs is easier to quantify than the eventual returns on massive AI capital expenditure, he said.

And the warnings from Apple's suppliers suggest that the cost pressure is unlikely to ease quickly. Last month, Samsung, a major supplier of RAM(DRAM), NAND flash storage and OLED displays for Apple, said that memory demand would far exceed supply as AI infrastructure continues to expand. Its memory division also reported a quarter-on-quarter increase in DRAM average prices during the third quarter, while NAND prices rose by high-teen percentages. Senior executive vice president of its memory business warned the supply crunch could extend to 2028 and beyond. Elsewhere, SK Hynix delivered a similar warning, saying it had secured long-term agreements covering more than Half of its production capacity for2027 and 2028. Micron, the third major supplier for Apple aside from Samsung and SK Hynix, is also benefiting from the surging demand for memory. Its latest published outlook warns that supply shortages in memory and storage could take “considerable time” to improve.

We have previously reported several signs of the pressure now confronting Apple. In January, the company was already scrambling to secure memory supplies amid strong demand, with then CEO Tim Cook warning that rising prices could have a greater impact on earnings. By June, rising component expenses had translated into higher retail prices. Apple increased prices across several Mac and iPad models, including a $400 increase to the starting price of the MacBook Pro, due to the RAM crisis. In July, Cook described the memory market as a “hundred-year-flood” and said Apple expected to pay more for memory in the September quarter. The company had also increased its inventory to $11.09 billion as it built a buffer against supply and price pressures.

Intel's shares plunged 30% on Friday, wiping out nearly $150 billion in market capitalization in a single day, the largest single-day loss in the company's history. The decline followed a disappointing financial outlook, largely driven by an anticipated drop of more than 20% in sales at Intel's client computing group (CCPG), the company's main cash cow. The declines quickly spread across the semiconductor industry. Although losses elsewhere were nowhere near as severe as Intel's, approximately $1 trillion in semiconductor market capitalization was wiped out on Friday, according to intraday estimates by Benzinga.

Fire sale

The PHLX Semiconductor Sector Index (SOX), which tracks the stock performance of 30 major semiconductor companies, is one of the key indicators of the industry's financial health. On Friday, the index plunged 9.14%, or 1,052.82 points, to 10,465.38 points as of 2:31 PM ET, as Intel's disappointing outlook triggered a massive selloff across the semiconductor industry.

Intel was by far the biggest loser, as other chip-related stocks, while Arm (-12.68%), AMD (-12.48%), Qualcomm (-12.14%), Applied Materials (-8.73%), ASML (-8.13%), Lam Research (-8%), Micron (-7.28%), and Broadcom (-7.1%) suffered considerably less. Then again, while Nvidia lost 5.66% of its market cap, this translates to around $270 billion, which is a massive amount of money. Paradoxically, Nvidia lost more market capitalization in absolute dollar terms than Intel.

In an unexpected turn of events, the downturn extended to server manufacturers and cloud giants too, as Dell (-12.67%), HPE (-8.10%), Supermicro (-6.96%), and Oracle (-6.88%) lost nearly 7% to 13% of their value. Reuters noted that investors increasingly question the scale of capital investments required to expand AI infrastructure as well as the sustainability of that spending.

Gloomy outlook

Intel’s client computing and physical AI group (CCPG) generated $8.1 billion in revenue during Q3 2026, down 5% year-over-year and 9% sequentially. Weaker demand and increased competition reduced client processor shipments by 14% year-over-year, which was partially offset by a 9% increase in average selling prices. CCPG accounted for nearly half of Intel's revenue of around $16.5 billion and the company's main source of cash. As strong demand for server processors only partially offsets declining client sales, Intel guided its Q4 revenue to be between $15 billion and $16 billion.

Intel expects the client market to deteriorate further in Q4 2026 and into 2027 as elevated memory prices and shortages of other components impact PC prices and demand. In general, Intel expects the client PC market to decline 10% in 2027 compared to 2026, which will inevitably have an effect on CCPG performance. While Intel expects its data center business to continue growing in 2027, industry-wide supply constraints will limit its ability to fully capitalize on demand.

2026 has been a mixed bag for Intel so far. On the one hand, the company achieved its highest market capitalization ever — $673.43 billion in mid-June — on growing sales of data center CPUs, strong sales of client products, and a return to profitability (if the $12.5 billion non-cash accounting charge associated with shares issued to the U.S. government is excluded). On the other hand, the company's capitalization has dropped by nearly 50% to $338.91 billion since mid-June (with $148.68 billion in just one day), as investors began to realize that the performance of Intel's CCPG heavily depends on memory prices, as customers and businesses tend not to like overpriced systems. Yet on the other hand, now that Intel sells fewer client processors, it can sell more expensive data center CPUs.

What about Apple?

Surprisingly, Apple (-1.66%), which only sells client products, was nearly unaffected by the performance of Intel's client computing group, as investors probably believe that the world's largest supplier of consumer electronics is more immune to memory prices than Intel. Ironically, if we count the A-series SoCs used in iPhones (which tend to outperform mainstream laptop CPUs), Apple sells substantially more processors than Intel, making Apple by far the world's largest supplier of microprocessors. Despite this, Apple's shares were largely unaffected by the semiconductor selloff.

There are several reasons for this distinction. Firstly, Apple's business model is fundamentally different. Intel sells processors to PC manufacturers, whereas Apple designs its own processors and sells them as part of finished products. Consequently, Apple captures revenue from the entire device rather than just the processor. Secondly, the two companies face different demand dynamics. Intel depends heavily on the PC market, while Apple's iPhone business follows the smartphone replacement cycle. Rising memory prices could affect both, but not necessarily to the same extent and at the same time.

There is an additional irony: Apple is also a major customer of TSMC, so its processor supply chain overlaps with those of AMD, Intel, and other semiconductor companies whose shares declined sharply.

While we should not directly compare Apple and Intel, the comparison illustrates that investors were not necessarily abandoning companies that design or use semiconductors, but were rather concerned about the profitability and growth prospects of these high-tech companies. And if we take a look at the situation from this point of view, the whole thing will start looking completely different.

Apple finally has touchscreen laptops. At an event in Los Angeles, the company detailed its new MacBook Pros featuring OLED XDR displays that you can control with your fingers.

The 14-inch and 16-inch MacBook Pros are available for pre-order now and will ship on Nov. 6. The 14-incher will start at $2,899, while the 16-incher will start at $3,399. The systems use Apple's existing M5 Pro and M5 Max chips, and go up to 16TB of storage and 128GB of memory.

These systems also feature the first redesigns for the MacBook Pro since 2021, when the M1 Pro and M1 Max launched in what has become a familiar design with a notch. These new models include the Dynamic Island around the webcam, similar to the iPhone, which can be used for real-time notifications. They also have light-up Apple logos, which we haven't seen on a MacBook Pro since 2016.

The PROLED XDR display comes in 14.2-and 16.3-inch sizes, with ProMotion allowing for refresh rates between 10 Hz and 120 Hz. Apple claims 2,000 nits of peak brightness and a 10,000,000:1 contrast ratio for intensely deep blacks.

The 14-inch MacBook Pro is 3 pounds, about half a pound less than the existing version, while the 16-inch MacBook Pro is 4 pounds, or 0.7 pounds lighter than it was before. Apple's bezels are also less noticeable than before. The hinge seems to be a point of pride (that's the second time this month, after the iPhone Duo). In this case, it's that the screen is less prone to wobble when you touch it.

The new laptops are only slightly thinner than their predecessor. Apple claims that the 16-inch model is "nearly a millimeter thinner" and the 14-incher is "half a millimeter thinner." Apple attributes those shavings to OLED's display stack.

Apple's own apps are already set to work with touch out of the box, and the company says many other apps will already work right away, including Notion, Canva, Zoom, Microsoft Word, Excel, PowerPoint, Adobe Lightroom, DaVinci Resolve, Blender, and more.

Apple already had the best speakers in a laptop, and now it's promising improvements: 50% deeper bass and new tweeters for clearer sound. There's even dedicated audio tuning for people who want to use the speakers with their laptop lid closed.

The significant change to the Mac is another in a flurry of launches for new CEO John Ternus, who also recently introduced the iPhone 18 Pro; the foldable iPhone Duo; Apple Watch Series 12 and Ultra 4; HomeView; and a new iPad Mini, also with an OLED panel.

The devices follow a long resistance to touchscreen laptops from Apple. Former CEO Steve Jobs, at an event introducing Mac OS X Lion in 2010, called laptops with touchscreens "ergonomically terrible", claiming that "after an extended period of time, your arm wants to fall off." (HomeView, Apple's smart home manager, has a vertical touchscreen and no one has made a peep about that. The iPad, too, has been vertical on Apple's keyboards, though some of those include touchpads as well).

The Mac resisted a touchscreen throughout former CEO Tim Cook's tenure, as well, though these new laptops definitely were developed under his leadership. In the last decade, touchscreens have become a common feature on laptops running Windows, and Microsoft's latest operating system, Windows 11, was designed to be slightly more touch-friendly than its predecessors.

Apple is updating two more computers to its M6 system-on-chip. The base 14-inch MacBook Pro and the iMac lineup are getting the company's latest processor ahead of tomorrow's "Pro and Tell" event in Los Angeles.

These devices are retaining the same designs that they've had for years. The computers will be available for pre-order on Wednesday at 10:10 a.m. PT (1:10 p.m. ET) and will ship on Nov. 6. The M6 in the MacBook Pro will be configured with 12 CPU cores and 12 GPU cores, alongside dual neural engines for certain on-device AI workloads. On the iMac, there will also be a cut-down option with a 10-core CPU and 10-core GPU.

The MacBook Pro will be configurable with up to 32GB of RAM and is being updated with Wi-Fi 7 and Bluetooth 6 using Apple's N1 networking chip. The iMac is also getting those networking benefits, as well as a 2.5 Gigabit Ethernet port. Like previous iMacs, the number of ports you get is tied to the chip, with two Thunderbolt 4 ports with the base chip or four with the option with more cores.

It's no surprise that Apple is pushing both of these as AI-forward machines, both with the new chips and the recently released Siri AI and Apple Intelligence features in macOS 27 Golden Gate.

The 14-inch MacBook Pro with M6 will start at $1,999 ($1,899 with education discount), while the iMac will begin at $1,499 ($1,449 with education discount) for the 10-core CPU/GPU configuration and $1,699 ($1,599 with education discount) for the full chip.

Both of these new systems are being announced the day before Apple is expected to release a MacBook with an OLED touchscreen. Yesterday, in a week of announcements, Apple debuted a redesign of the iPad Mini with a thinner body and an OLED display.

Apple internally uses a system known as “Minions,” a “virtual GPU cloud” made up of idle Macs, to help manage workloads, as described in a recent arXiv paper published by its researchers. The system helps take stalled AI jobs, some that once had 10-plus-hour company GPU queues, and distributes them to Macs on the network, with resulting wait times down to as little as minutes. A request from a user on the company’s Slack can get the ball rolling, although the researchers observed a common concern about teams’ data running on devices outside their group.

The paper, also shared on Apple’s Machine Learning Research site, describes the system as having a “Minion Server,” a “Minion Executor,” and a “Minion Watchdog.” The system relies specifically on Apple Silicon for AI inference, the only type of work it does. There are two kinds of jobs: embeddings with open-source Hugging Face transformer models, and text generation from large language models through Ollama, which is installed on each participating Mac.

The Minion Server works by splitting submitted tasks into subtasks, stored in a PostgreSQL database. Then, on each Mac, a Minion Executor is run to pull the next subtask over REST APIs; there is no central scheduler in this “pull-based execution model.” The Mac reads inputs from Amazon S3, grabs the model if needed, runs the subtask on its own GPU, then returns the result to S3.

Failures are caught by the Minion Watchdog, which tracks the heartbeats that Macs running work send every five minutes. Unfinished subtasks end up back in the queue when, for example, a Mac goes to sleep, drops off the network, powers off, or a user stops the executor to get control of the machine back. The owner can determine which jobs the Mac takes first, such as preferring one team’s or project’s jobs with “dynamic task binding,” or can rely on the default first-in, first-out scheme.

Our understanding is that each Mac runs a whole model on its own piece of work, and the paper does not suggest that any one model is split across Macs. Running a model entirely on one Mac’s hardware with unified memory should come as no surprise to OpenClaw enthusiasts who practice the same in their own home labs. The volunteer computing angle is also not new, reminiscent of SETI@home’s and Folding@home’s model of distributing small chunks to idle machines, based on the paper’s described architecture.

Aside from the reduction in GPU queue times, the paper also cites that requests with Minions do not require new GPU requests, the infrastructure cost is zeroed, and the scaling effort goes from manual to organic. This does not factor in the electricity costs or hardware wear.

Participation is handled informally, and team members may, for example, “send Slack messages requesting others to temporarily run the Minion Executor and target a specific project or task.” Confidentiality remains a concern even for internal work, which Apple aims to mitigate with team- and organization-scoped deployments with their own databases and separate pools of Macs.

The new paper comes at a time when compute is increasingly a bottleneck for AI-related workloads. Apple acknowledges this in its latest quarterly filing in July, saying that it “depends on access to sufficient computing resources” in an ecosystem with “constrained supply, extended lead times, and increasing costs.” The risk that it may be unable to “secure sufficient capacity,” even with its own data centers and third-party cloud providers, appears to make the Minions approach more practical.

Apple’s Private Cloud Compute, the server side of Apple Intelligence for more complex requests, now also runs on Nvidia GPUs in Google Cloud, first announced at WWDC in June. That came after sudden and high Mac demand for local AI, with then-CEO Tim Cook saying in April that the Mac mini and Mac Studio supply may take months to catch up. Device supply discussed in the July earnings call also mentioned advanced-node constraints and increasing memory costs.

The authors say that, in future work, Minions will extend to distributed model training. More broadly, they write that if idle Macs across organizations “or potentially across the global community” can be “securely federated,” they could form a shared GPU pool. For now, the Macs act as overflow AI capacity only inside Apple.

Apple's entire range of computer products received an unfortunate price hike earlier in the year, an unavoidable side effect of the AI boom and scarcity of affordable memory modules. Now's a great opportunity to pick up a MacBook Air, MacBook Pro, or Mac Mini for a slightly reduced price, with savings up to $600 on some models.

Luckily for shoppers and deal hunters, we're in the beginning throes of Amazon's Prime Big Deal Days sales week, and discounts are live on Amazon for a host of 2026/2025 Apple MacBook models and Mac Mini computers.

● Check out these Apple deals at Amazon

These deals may not be around for long. With prices ever rising on anything that hosts speedy memory or storage, the cost of Apple products is sure to rise again in the near future. There's a good selection of configurations to suit your needs, with various amounts of either memory or storage offered in the different configurations.

Although not discounted, the Apple MacBook Neo is one of the best budget laptops that you can get your hands on. Definitely worth keeping your eye out for any future deals on the Neo as the sales continue throughout the week.

Apple MacBook Neo Amazon Prime Big Deal Days Deals

Apple MacBook Air Amazon Prime Big Deal Days Deals

Apple MacBook Pro Amazon Prime Big Deal Days Deals

Apple Mac Mini Amazon Prime Big Deal Days Deals

More Amazon Prime Big Deal Days Tech Deals

Best Tech and PC deals | Best gaming PC deals | Best RAM combo deals | Best 3D printer deals | Best RAM deals | Best gaming laptop deals | Best monitor deals | Best Wi-Fi Router deals | Best GPU deals | Best SSD deals | Best hard drive HDD deals | Best CPU deals | Best gaming chair deals | Best PC building tool deals | Best PC peripherals deals | Best filament and resin deals | Best motherboard deals | Best CPU cooler deals | Best PC case deals | Best Dell and Alienware deals | Best USB charger deals | Best gaming and productivity laptop deals under $1,000 | Best laptop PC deals

Also, you can join the Tom's Hardware deals Discord for up-to-the-minute hardware deals.

When Nuvacore, a new CPU startup founded by legendary CPU and system architects Gerard Williams, John Bruno, and Ram Srinivasan, emerged from stealth earlier this year, it promised to "rewrite the rules of silicon," but did not say how it planned to do so. On Thursday, the company finally broke its silence and revealed that it is taking an unusual development approach that allows it to design a significant portion of its CPU core IP before committing to a particular instruction set architecture (ISA).

Traditionally, CPU developers start development of a new processor project with a particular ISA — such as Arm, RISC-V, or x86 — already selected. Although modern high-performance processors translate instructions into internal micro-operations and therefore separate the ISA from much of the underlying execution pipelines, the instruction set still influences numerous architectural decisions, including instruction decoding, register handling, memory ordering, and, of course, software compatibility.

Nuvacore says its Core First approach reverses at least part of this process. Instead of designing WarpCore around the requirements and limitations of a predetermined ISA, the company's engineers are developing a significant portion of the foundational CPU IP independent of a particular ISA. To a large degree, one can compare such an approach to the development of a car around its engine and not trying to fit an engine into an already developed platform.

"Rather than beginning with the constraints of an existing architecture and iterating from there, we are starting with the core itself," the company said in its statement.

As a result, the team may be able to better focus on the core's microarchitecture and optimize it for performance, power efficiency, and sustained workloads typical of modern data centers and AI infrastructure, the target application for the company. Nuvacore intends to select an instruction set later based on what works best for the systems that it and its partners plan to build.

There are obvious limits to how ISA-independent a CPU design can be. Eventually, WarpCore will need front-end logic designed around its chosen instruction set, and those architectural requirements will inevitably affect other parts of the processor. Nevertheless, substantial pieces of a modern CPU — such as execution units, branch-prediction units, data paths, caches, and portions of the memory subsystem — can be developed and evaluated before every ISA-related implementation decision is finalized.

In addition to not disclosing its ISA, which is arguably the biggest unanswered question surrounding the project, Nuvacore also has not disclosed the project's schedule, configuration, manufacturing process, power, or performance targets, though given the lack of the ISA commitment, one could say that the remaining details are not that important.

For now, the only thing that Nuvacore discloses about WarpCore is that it is "a new class of general-purpose CPU designed specifically for the sustained performance and power-efficiency requirements of AI infrastructure and contemporary data centers."

AMD's ambidextrous strategy returns?

Nuvacore's approach is not entirely without precedent. Under CEOs Rory Read and Lisa Su, AMD spent part of the 2010s pursuing an "ambidextrous" strategy intended to let it address x86 and Arm customers by sharing substantial processor and platform IP.

Arguably the most detailed part of the initiative was AMD's SkyBridge platform, announced in 2014. Under SkyBridge, AMD envisioned pin-compatible 64-bit Arm and x86 systems-on-chips that were to use AMD's own Puma+ x86 cores and Arm's Cortex-A57 cores with common platform infrastructure. Separately, AMD developed its custom 64-bit Arm K12 core under Jim Keller that was supposed to share microarchitecture features with Zen. Ultimately, due to lack of resources, SkyBridge was ultimately scrapped, while K12 never became a commercial product as AMD only pursued x86 Zen cores. The difference is, of course, that Nuvacore appears to be taking the idea further by deliberately postponing the ISA decision for WarpCore.

Since Nuvacore is still just a startup, we can only wonder whether its business plan includes marketing of CPUs at all. Nuvia, which was also founded by Williams and Bruno along with Manu Gulati, was acquired by Qualcomm and ultimately shaped the company's Snapdragon X Elite SoCs. In that light, its approach could give the company considerable commercial flexibility.

By developing much of its WarpCore design without tying it to a particular ISA, the company could potentially adapt the technology to the requirements of a future customer or partner. An Arm licensee could use the underlying core technology for an Arm implementation; another customer could choose RISC-V or another ISA. An established x86 vendor could potentially acquire or license the underlying microarchitecture and create its own x86 CPUs if it needs to.

Nuvacore, of course, has not indicated that such a transaction is its objective, but keeping WarpCore ISA-agnostic for as long as possible potentially broadens the number of companies to which its CPU technology could be valuable.

All in all, the Core First announcement tells us considerably more about how Nuvacore intends to develop WarpCore than what the processor itself will eventually look like. But if the company really does keep its ISA options open until relatively late in development, WarpCore represents an unusual attempt to separate development of a high-performance CPU microarchitecture from one of the decisions that normally defines a processor project from its earliest stages.

Apple’s latest iPhone 18 Pro series went on sale last week, and the company’s decision to use QLC-based NAND for higher-capacity storage variants appears to come with a trade-off. Bilibili channel Homolab recently tested a 1TB iPhone 18 Pro Max and found that its storage performance can drop significantly under sustained heavy workloads. Since QLC stores four bits of data per memory cell when compared to three bits with TLC, it allows manufacturers to offer higher capacities using the same physical space. However, this higher storage density comes at the cost of slower performance, particularly when handling large amounts of data.

Homolab compared the 1TB QLC-based iPhone 18 Pro Max with a 512GB TLC-based iPhone 18 Pro. While both storage types perform fairly similarly in 4K read tests, the QLC model falls behind by 38% in a low-queue-depth mixed workload, scoring 8,168 compared to 11,285 for the TLC model. Under heavier workloads, the gap narrows, although the TLC storage model still maintains a 12% advantage.

The drop in performance is even more apparent when it comes to sustained write tests. The iPhone 18 Pro Max uses a fast SLC cache that can deliver speeds of up to 3,000 MB/s. Once that cache is exhausted, a secondary TLC-based buffer takes over, averaging around 578 MB/s. When both cache buffers are exhausted, the phone has to write directly to the underlying QLC NAND. At that point, sustained write performance can fall to 79.4 MB/s, with speeds occasionally dropping as low as 25.6 MB/s.

That is significantly slow and comparable to the performance of many budget microSD cards (microSD Express cards are much faster). Things look even worse once the storage starts filling up. At 60% capacity, the available SLC cache drops from 250GB to just 58GB. This also causes the secondary buffer to slow down to 396 MB/s, while direct QLC writes average just 45 MB/s and occasionally fall to as low as 1.1 MB/s.

It’s important to note that these are synthetic benchmarks and are meant to represent a worse case scenario. Users should not experience any performance issues during typical tasks such as web browsing, messaging, gaming, or taking photos. Any slowdown is likely to be noticeable only during specific workloads, such as 4K or ProRes recording, large file transfers, or data backup restores. Having said that, testing by Homolab does point to the trade-off involved in using QLC NAND for a high-capacity flagship smartphone, particularly for users who regularly deal with storage-intensive tasks.

It’s been nearly two years since Apple updated the Mac mini. And for better or worse, things have changed a lot during that time. The last Mac mini with an M4 processor went on sale in early November 2024, with a starting price of $599. The latest Mac mini that we’re testing today starts at $899, which we can directly blame on the rising costs of components due to the voracious demand for everything related to artificial intelligence (AI).

A $300 price hike in less than two years is unprecedented, but it’s a sign of the times, and we have to factor those realities when evaluating any product today. So with that context in mind, we’ll take a look at Apple’s newest Mac mini, which looks very familiar on the outside, but has undergone some internal hardware upgrades, to deliver some significant performance improvements over its predecessor.

Design of the Apple Mac mini (M6)

The Mac mini hasn’t changed on the design front, and that’s mostly a good thing. I’m always impressed with not only the minimalist style of this small form-factor computer, but also its tiny size. It has a 5-inch x 5-inch footprint, while standing just 2 inches tall. My only gripe is the power button, which remains stubbornly on the bottom of the system (back left corner). You’ll need to lift the system any time you want to turn it on (or force it to shut down in the event of a crash), and you’ll need to double-tap the power button when setting up Touch ID with an Apple Magic Keyboard for the first time. However, this is just a minor annoyance, as interacting with the power button on a modern Mac is a rarity.

Despite its trim dimensions, you do get a fairly decent number of ports on the exterior of the Mac mini. However, let’s get this out of the way first: Apple doesn’t give you any USB-A ports on the Mac mini, so keep that in mind before purchasing. So, if you have USB-A devices around, you’ll either need to keep some cheap USB-A to USB-C adapters on hand, or invest in a USB-C or Thunderbolt dock for connecting legacy devices.

Instead, our Mac mini review unit features two USB-C ports (10 Gbps) and a 3.5 mm headphone jack up front, along with three Thunderbolt 4 ports on the rear (M5 Pro systems swap the Thunderbolt 4 ports for Thunderbolt 5). You’ll also find a 2.5 GbE port (up from 1 GbE), an HDMI port, and the power port on the back. We should note that while our review unit was specced with the 2.5 GbE LAN, you can individually option a 10 GbE port for $100 extra. A 10 GbE USB-C adapter would set you back around $100, so Apple’s upgrade is surprisingly reasonably priced, and you don’t lose a port in the process.

When it comes to wireless connectivity, the Mac mini uses Apple’s new in-house N1 wireless chip instead of relying on a third-party design. It provides tri-band Wi-Fi 7 and Bluetooth 6 support, which is an upgrade from the Wi-Fi 6E and Bluetooth 5.3 on the previous generation.

One thing to keep in mind with the Mac mini is that what you see is what you get when you purchase one. There are no internal upgrades available for the system once you’ve purchased it, so choose wisely before you click the buy button. Base systems come with an M6 chip (12-core CPU, 12-core GPU), 16GB of RAM, and a 256GB SSD. Our review unit features the M6 chip, 24GB of RAM, and a 512GB SSD. If you want to add storage post-purchase, we’d suggest taking a look at one of the best external SSDs.

Apple Mac mini (M6) Specifications

Productivity Performance on the Apple Mac mini (M6)

Although the Mac mini is available with a more potent M5 Pro processor, our review unit came with the base M6 processor (12 CPU cores, 12 GPU cores). It also incorporates 24GB of LPDDR5X memory and a 512GB SSD.

This is the first system we’ve tested with the new 2-nanometer M6 processor. So, we were eager to see how it would perform against not only the M4 Pro in the previous Mac mini, but also the M5, which can be found in the current-generation MacBook Air and base MacBook Pro.

The M6 chip in the Mac mini had no trouble dispatching the M4, M4 Pro, and M5-equipped systems in the Geekbench synthetic CPU benchmark. In fact, it was 21 percent faster than the last M4 Pro Mac Mini in the single-core test. It also completely obliterated the M4 (10-core CPU, 10-core GPU) version of the Mac mini. The advantage versus the M5 system was lessened, but it was still a clear victory for the M6. The M6 actually slightly trailed the M4 Pro Mac mini (21,045 versus 22,661) in the multi-core test, which wasn’t too surprising given that the 12-core M6 was going up against a 16-core M4 Pro. But it was still a strong showing for the M6, given the architectural improvements that allowed it to get that close.

We had some trouble obtaining consistent performance from our usual 25GB file transfer test, so we fell back to the Blackmagic Disk Speed Test for benchmarking the 512GB SSD. Needless to say, it presented a marked improvement over the preceding M4 Mac mini.

Storage performance was comparable to the M5 MacBook Pro on sequential writes, but slightly behind for sequential reads. It also slightly outperformed the M4 Pro Mac mini in writes, but was over 1,000 MBps ahead in reads. However, the M6 Mac Mini nearly doubled the write performance and more than doubled the read performance of the M4 Mac mini.

The similar multi-core performance in Geekbench was mirrored in our HandBrake test, during which we transcoded a 4K video file to 1080p. The M6 Mac mini completed the task in 2 minutes and 46 seconds, while the M4 Pro version was just one second behind. The M6 Mac mini was nearly 2 minutes faster than its M4 counterpart.

We ran Cinebench 2026 through 10 runs with the Mac mini, keeping track of its scores along the way. The Mac mini started off with a score of 5403, which rose to 5557 for the second run. For the remaining eight runs, the Mac mini was highly consistent, hovering between 5530 and 5554.

The M6 used in the Mac mini isn't yet fully supported by TG Pro, so we could only use the average of the available temperature sensors (power management sensors, power supply, and SSD). At idle, the fan spun at just around 1,000 revolutions per minute (RPM), with an average system temperature of 44 degrees Celsius. However, when running Cinebench 2026, the average system temperature rose to 64 degrees C, while the fan settled at around 2615 RPM.

Gaming and Graphics on the Apple Mac mini (M6)

Gaming shouldn’t be the primary goal with the Mac mini, but it can do some light lifting. I played Cyberpunk 2077 for a bit and benchmarked it. Although I was provided an Apple Studio Display (5K resolution) to test with the Mac mini, I ran benchmarks at 2560 x 1440 and 1920 x 1080, because the native 5K is a bridge too far for this game, given the hardware.

Using the "For this Mac" global setting, the Mac mini achieved 37.33 frames per second (fps) at 1440p and 58.11 fps at 1080p. With frame generation turned on, those numbers jumped to 62.61 fps and 96.24 fps, respectively.

If you don't mind turning the detail settings down a bit, you can achieve playable frame rates with the Mac mini. However, moving the detail slider to Ray Tracing (Medium) dropped frame rates into the 20 fps range even with frame generation turned on at 1440p.

macOS Golden Gate and Apple Intelligence

The Mac mini ships with macOS 27 Golden Gate, which was first made available to the public last week. macOS 27 is not a complete rethink of the desktop operating system formula for Apple, but it does provide a number of quality-of-life improvements for users.

One of the headlining features is new and improved Siri AI. While Siri was one of the first digital assistants to debut on a smartphone, it hasn't exactly been "smart," at least compared to its contemporaries. Siri AI now digs its fingers deeper into the operating system, allowing it to glean contextual data from your photos, mail, messages, and documents. You can access Siri AI from Spotlight (search) or using the dedicated Siri AI app. Siri AI is also vastly more useful, as it will more consistently answer general questions asked, rather than throw up its digital hands and say "Here's what I found on the web."

Apple's Shortcuts app is an incredibly powerful tool that has been available for years across the Mac, iPhone, and iPad product families. However, the granular control and seemingly limitless options can make creating a new shortcut daunting for newbies. So with macOS 27, you can now create a new Shortcut using natural language. For example, you could input, "Create a shortcut that sets the AC to 74 degrees when I leave work, and let my wife know what time I'll arrive at home," and it will take care of the rest.

Liquid Glass was one of the more controversial additions to macOS 26, with some people not liking the translucent makeover. MacOS 27 gives users more control over Liquid Glass, with sliders to adjust the intensity of the transparency effect.

Apple has also introduced plenty of system-level optimizations to improve performance across the board. For example, Apple says that AirDrop is up to 80 percent faster in macOS 27 than macOS 26, and a lot more reliable. Apps also open faster – any lagginess or stuttering that was perceptible in macOS 26 has been eliminated with this newest release. The best comparison I can make for the macOS 26 to macOS 27 upgrade is the fine-tuning that Apple did going from Mac OS X 10.5 Leopard to Mac OS X 10.6 Snow Leopard. There weren’t any major feature additions with that latter release, but there were a plethora of under-the-hood optimizations that made it much faster and more stable.

In addition, macOS 27 doesn’t install the Rosetta x86 translation software by default; you’ll need to download it when you first run a legacy x86 app. However, starting next year with macOS 28, you won’t be able to run apps that aren’t Apple Silicon-native, so Rosetta will be completely off the table.

Apple Mac mini (M6) Configurations and Warranty

The base Mac Mini comes with an M6 processor, 16GB of RAM, and a 256GB SSD for $899. Before Apple's June 2026 price increases, the base Mac mini with M4, 16GB of RAM, and a 256GB SSD cost $599. After the price increases, that $599 jumped to $799. Now, we're looking at another $100 increase for the base Mac Mini. There’s one standalone upgrade option: $100 extra gets you a 10 GbE port instead of the standard 2.5 GbE.

Apple also allows you to upgrade to 24GB or 32GB of RAM, along with 512GB, 1TB, and 2TB SSD options. A fully-loaded M6 Mac mini with 32GB of RAM, a 2TB SSD, and 10 GbE costs $2,399.

For people who require a bit more processing power, perhaps for AI tasks (as Apple seems so inclined to believe), the base M5 Pro Mac Mini (15-core CPU, 16-core GPU), with 24GB of RAM and a 512GB SSD, costs $1,699. However, you can go bonkers with the options, topping out with an M5 Pro (18-core CPU, 20-core GPU), 64GB of RAM, an 8TB SSD, and a 10 GbE adapter for a whopping $6,599. At that price, it’s probably worth looking at the Mac Studio.

All Mac minis come with a 1-year manufacturer warranty.

Bottom Line

The base M6 chip provides better single-core performance and nearly matches the multi-core performance of the optional M4 Pro in the previous-generation Mac mini. Compared ot the standard M4 Mac mini, the new M6 version blows it away in CPU and SSD performance.

These are the kinds of generational improvements we’ve come to expect with Apple Silicon, which seems to be firing on all cylinders these days. The new Mac mini also gets other, less-highlighted improvements, like faster standard wired networking (2.5 GbE versus 1 GbE) and Wi-Fi 7/Bluetooth 6.0 instead of Wi-Fi 6E/Bluetooth 5.4.

However, the elephant in the room is pricing for the Mac mini. Earlier this year, you could grab a base Mac mini for $599 – although availability was spotty after Q1, thanks to OpenClaw and anticipated price hikes. More importantly, over the years, going from one Mac mini generation to the next typically yielded performance benefits with no (or minimal) price increase. The AI revolution has rewritten the hardware rulebook, so we’re now faced with a new generation that is $300 more expensive than the previous model ($200 June price increase + another $100 increase for this new generation).

So in a matter of months, what was for years a screaming bargain for anyone looking for a cheap entry point into the Mac desktop ecosystem has gotten a lot more expensive. 16GB of memory and a 256GB hard drive are easier to justify at $599, but it’s a tougher pill to swallow at $899, and especially so with our review unit’s as-tested price of $1,399.

Earlier this year, Apple discontinued the Mac Pro, leaving the Mac Studio as its top-end desktop. With the new M5 Ultra chip, Apple's relatively diminutive system is more than ready to take on the mantle.

While you might not have a ton of PCIe connections and both RAM and storage are set, the Studio offers a ton of ports for accessories, monitors, and peripherals, memory options up to 256GB (with 512GB coming later), and a clean, quiet design.

Starting with an M5 Max at $2,499, the Mac Studio could be poised as a creator system for video editors, photographers, or game designers. At $12,299 as tested with the top-end M5 Ultra, 256GB of unified memory, and 4TB of storage, it effectively becomes an enterprise device, possibly a remote server, that could handle any task with aplomb, including local AI.

Design of the Mac Studio (2026)

Apple hasn't made any changes to the Mac Studio's design. It's a silver, squared-off block that's 7.7 x 7.7 inches on your desk that stands 3.7 inches tall. The aluminum chassis has a shiny Apple logo on the top. The M5 Ultra version we tested weighs 8 pounds, a 2-pound increase over the M5 Max variant. This is largely because the M5 Max version uses a cooler with a thin aluminum stack and a copper heat pipe, while the M5 Ultra uses a heavier copper fin stack and a copper vapor chamber.

The back of the system is also home to the power button, as well as a bunch of air holes for exhausting heat.

The air comes in through more holes in a small stand built into the bottom of the system. (There's also a Kensington lock slot, though its placement on the bottom necessitates a special adapter.)

Mac Studio (2026) Specifications

Ports and Upgradeability on the Mac Studio (2026)

On the front, there are two Thunderbolt 5 (USB Type-C) ports along with an SDXC card slot. The rest of the ports are on the rear, including another four Thunderbolt 5 ports, two USB Type-A ports, HDMI 2.1, a 10Gb Ethernet jack, and a 3.5 mm headphone jack.

There is no upgrading the Mac Studio. The RAM is soldered to the board, so be sure to get what you need when you're first buying the system, and Apple doesn't sell replacement SSDs (Some third parties may sell reverse-engineered NAND, but it will certainly void your warranty, and Apple doesn't support it. You're best off using external storage)

M5 Ultra

If you pony up for the most expensive Mac Studio configurations, you'll get Apple's most powerful chip. This is the only machine that Apple currently puts that chip inside.

M5 Ultra is Apple's first quad-die chip, connecting a pair of dual-die M5 Max SOCs over its UltraFusion connection. Apple claims that UltraFusion increases bandwidth between the dies to over 4.4 TB/s. Memory bandwidth is up to 1.2TB/s, which lets users store and run huge LLMs entirely on device.

Each of the 80 GPU cores features a neural accelerator, and there's a 32-core neural engine. There's now a media engine with hardware-enabled H.264, HEVC, hardware-accelerated AV1 decode, and four ProRes encode and decode engines.

Cinebench estimates that the M5 Ultra runs single-core at 4.6 GHz and multi-core at 4.3 GHz.

Productivity Performance on the Mac Studio

The Mac Studio is an absolute workhorse. The system stayed largely quiet (though the fan can make itself known) as it powered through our benchmarks. Anything you might want to do on a Mac, you can do on this one.

In fact, while we ran our basic desktop suite on this, we also ran some extra tests, as we don't test many workstations. Instead of pitting this system against other desktops, we pitted it against other workstations to get a better idea of how the M5 Ultra performs. While that's a bit antithetical to the way Apple designs its products (it doesn't sell chips separately; it designs them as part of its computers), it does give us a reference point to the chips powering the Windows and Linux workstations.

On Geekbench 7, the Mac Studio was the clear winner, with the ARM-based chip leading the pack in the bursty workload with a single-core score of 3,770 and a multi-core score of 52,293, despite having fewer cores than the competitors.

The Mac continued its lead on Handbrake, transcoding a 4K video to 1080p in 1 minute and 4 seconds. The next closest, the AMD Ryzen Threadripper 9980X, took 8 seconds longer.

On Blender's benchmark testing CPU-based rendering, the Studio with M5 Ultra fell behind the previous- and current-gen Threadrippers (7980X and 9980X, each with 64-cores), as well as the 56-core Intel Xeon w-3495X.

The Mac Studio with M5 Ultra copied 25GB of files at a rate of 2,983.08 MBps, surpassing the speeds of the M3 Ultra and M4 Max from the last generation.

The Mac Studio was consistent over our 10-run Cinebench 2026 stress test. Its lowest score, 17,825, was actually the first run. When the cooling kicked in, the rest of the results were slightly higher. The sixth run was the highest at 18,069, but it was so similar to the other runs that it's within the margin of error.

During the test, the CPU measured an average of 78.29 degrees Celsius.

Local AI on the Mac Studio (2026)

The M5 Ultra's GPU cores feature neural accelerators, granting it some extra matrix math horsepower, and Apple has boosted the chip's memory bandwidth to 1.2 TB/s, double that of the M4 Max we previously tested for AI.

The M5 Ultra's prompt processing speeds are even faster than Nvidia's DGX Spark's, and its tokens-per-second throughput is double that of the M4 Max and almost four times higher than the DGX Spark across the board.

Our tests, here using Qwen 3.8-27B-Q4_K_M, a dense model that relies heavily on memory bandwidth, are largely aimed at GPU performance in isolation. But combined with the blazing-fast single- and multi-threaded performance of the 36-core CPU, the Mac Studio with M5 Ultra looks to be ready for anything one might want to do with local or agentic AI workloads — and that's before someone tries to put it into an MLX cluster.

In addition to its raw performance, our M5 Ultra sample's 256GB of memory means you can fit very large, intelligent single models or a range of different models onto a single node. I spent time running models far larger than our tests with Hermes and LM Studio with strong performance. If you are interested in local AI, this will definitely do the job.

Gaming and Graphics on the Mac Studio (2026)

While the M5 Ultra and its 80-core GPU are more than capable of gaming, the Mac Studio isn't a gaming-first rig. For the price, systems with Nvidia's high-end graphics cards are far more competitive.

To stress the Mac Studio, I booted up InZoi, a life simulator that isn't graphically imposing but is hard on the CPU and memory. I set both the simulation performance and graphics to ultra settings, and set the game to 1080p, but allowed for upscaling with MetalFX's Performance mode, which upscaled from 960 x 540. The game ran between 45 and 65 frames per second, dropping to the lower end in outdoor environments with lots of other characters.

To compare more graphically demanding games, I picked out a few gaming PCs we've tested in the last year and pitted them against each other with the data from the Cyberpunk 2077 benchmark with the Ray Tracing Ultra preset, which we use for those kinds of rigs.

At 1080p, the Mac Studio ran the game at 66 FPS, falling behind systems like the HP Omen Max 45L and Maingear MG-1 with Nvidia's top-end GeForce RTX 5090. It was much more in line with the Acer Nitro 65, with an RTX 5070. At 4K, the Mac Studio didn't reach playable levels, though the RTX 5090 systems both reached 59 FPS.

Using Cyberpunk's "For This Mac" preset set the resolution to 1440p, turned off ray tracing, and mostly stuck to high settings and effects options.

On 3DMark's Steel Nomad graphics test, the Mac Studio still fell far behind the RTX 5090 system, but outperformed the RTX 5070-based Acer desktop.

Software and Warranty on the Mac Studio (2026)

The Mac Studio comes with macOS 27 Golden Gate. This is a great update over macOS 26 Tahoe, refining the Liquid Glass aesthetic and making it more customizable (I actually kind of like it in this iteration), and yes, thank heavens, consistent window shapes. It also adds the new Siri AI (be sure to get on the waitlist). The new Siri is far superior to the one you're used to.

Golden Gate also adds AI throughout apps, including more photo features, including a better Clean up option and spatial reframing, which lets you move a camera after you take a picture (the latter is kind of surreal). Better indexing makes it easier to search messages, emails, and calendar entries, and now you can use natural language to create automations in the Shortcuts app or come up with your own Safari extensions.

There's no bloatware, as usual. Apple's software suite includes Safari, Messages, Photos, Maps, Notes, and Reminders. Its services, including News and Apple TV, also have applications. Pages, Keynote, and Numbers are still in the dock by default, but they're the Creator Studio subscription versions, which require a subscription to get full features, including AI functionality, royalty-free graphics, and additional templates.

People using the Mac Studio as a server running AI models may not need all of those programs. For the first time in a while, I found myself wishing for something akin to macOS Server, but for better remote monitoring. I would also like to see an option for a system of this class to come with pared-down built-in apps, perhaps limiting it to the browser, App Store, Terminal, Activity Monitor, and other essentials and letting system admins decide what else needs or deserves to be there. A video editor might want to use Messages on the side. But you don't need that for a cluster running a huge model.

Apple sells the Mac Studio with a one-year warranty. Further repairs and support can be added as part of AppleCare+ ($6.49 monthly or $64.99 annually), or as part of AppleCare One or AppleCare Family.

Mac Studio (2026) Configurations

Our Mac Studio review configuration is clearly an enterprise machine. With the top-end M5 Ultra (36-core CPU, 80-core GPU), along with 256GB of RAM (currently the top-end option, with a 512GB configuration coming in late October), and a roomy 4TB SSD, it costs $12,299.00, but that's not stopping anyone: As of this writing, it's backordered between 16 and 18 weeks.

But there are a lot of steps below that. The base model is $2,499 with the M5 Max (18-core CPU, 32-core GPU), 36GB of memory, and 512GB of storage. There are two versions of each: the M5 Max and M5 Ultra. Going from a 32-core GPU to a 40-core GPU on the M5 Max costs $300.

The cheapest M5 Ultra system is $5,499 with a 30-core CPU and 64-core GPU, along with 96GB of memory and 1TB of SSD storage. Bumping up to the M5 Ultra we tested increases the price by $1,300. Moving from 96GB of RAM to 256GB adds $4,000. Bumping up from a 1TB SSD to 2TB is $500, while our 4TB SSD is a $1,500 increase.

The M5 Max configurations have slight port differences on the front of the desktop. The USB-C ports on those are 10Gb/s ports, while on the M5 Ultra, you get Thunderbolt 5 for up to 120Gb/s.

Bottom Line

At $12,299, the Mac Studio with M5 Ultra is effectively an enterprise computer, capable of difficult workloads like rendering and running AI models that are as close to frontier-grade as you're going to get right now on edge computing.

Much of that price also comes from RAM and storage. Given the fact that you can't really upgrade this computer, businesses and people using this for their job may see the extra cost as an investment. Really, its biggest weakness is in gaming, for professionals who may want to do that on the side. Discrete GPUs still win there.

Given the fact that an AMD Ryzen Threadripper 9980X, a 64-core monster that beats this CPU in Blender, is priced around $5,000 on its own before memory, storage, power supplies, or any other components, it may not actually be that crazy in this market, and the Mac Studio can be had in a much smaller case. (For wealthy homelabbers, businesses, and server farms, third-party accessory makers already make rack mounts to fit the Mac Studio; its shape hasn't changed since its last redesign).

For local AI, the M5 Ultra is seriously impressive, though that may still be a niche use case for some. But in a computer that can reach over $12,000, niche is the name of the game.

This all makes the M5 Ultra version of the Mac Studio an impressive package and further proof of the power and performance Apple can draw out of its own silicon in concert with its own software.

Apple is reportedly developing AI servers based on its own M-series processors and is evaluating NVLink Fusion technology for interconnects, according to The Information. The machines are expected to use M8 Ultra processors and arrive in 2029, the report claims.

For now, the usage of the NVLink Fusion platform is not formalized and has not been confirmed by either Apple or Nvidia — but an Apple decision to use it instead of competing solutions may have significantly broader market implications than just Apple using Nvidia hardware.

Apple looking for fast interconnects

Apple is reportedly considering at least two server configurations: a smaller machine equipped with two M8 Ultra processors and a higher-end version featuring four M8 Ultra system-on-chips. Although Apple has its own UltraFusion technology for stitching two high-end SoCs together seamlessly, it looks like the company does not have a proper solution for scale-up and scale-out connectivity of its processors, which is where Nvidia's NVLink Fusion comes into play. Apparently, Apple wants to use NVLink infrastructure, which includes not only an interconnection protocol, but also switches, chiplets that add NVLink connectivity, and a software stack, for its servers. The project was reportedly initiated around a year ago and was backed by John Ternus while he headed Apple's hardware engineering organization.

Apple already builds custom servers for Private Cloud Compute, which handle AI workloads too demanding for local execution on iPhones and Macs, The Information claims. Most of these machines use Apple's internally developed connectivity technologies, which are reportedly too slow and costly for large-scale commercial deployments, which is why Apple is looking elsewhere.

More than NVLink?

The Information specifically mentions Apple's need for connectivity technology suitable for large-scale deployments, although it does not explain exactly what this means architecturally. If the publication is referring to connecting multiple servers into larger clusters, this would normally be the job of scale-out technologies such as Ethernet or InfiniBand, rather than a scale-up fabric such as NVLink. Nvidia originally developed its NVLink fabric technology to scale-up performance of its accelerators, so the technology is optimized for accelerator-to-accelerator connectivity and enables a rack of Nvidia GPUs to function as a tightly coupled compute domain. There is a different implementation called NVLink-C2C, which is a coherent chip-to-chip interface for connecting CPUs to accelerators and CPUs to CPUs

Meanwhile, modern Apple M Pro and M Ultra processors are system-in-packages consisting of a CPU chiplet and a GPU/neural engine chiplet, which are stitched together using TSMC's SoIC-mH technology. If Apple continues to use this architecture (very likely), an M8 Ultra processor can be considered as a CPU and an accelerator. However, this raises the question of how Apple intends to connect M8 Ultra processors to NVLink and which components of the SiP would participate in the NVLink domain. One possibility is that Apple could expose the accelerator portion of M8 Ultra to NVLink through an NVLink Fusion chiplet, which effectively means it will treat it as an accelerator for a scale-up domain. Another possibility is that Apple is developing a different accelerator architecture for its servers, perhaps by simply placing the GPU/NPU chiplet onto a separate substrate/interposer and equipping it with its own memory, though there is currently no evidence that confirms such a design for a chip that is years away.

Another thing to keep in mind is that Apple is a member of the UALink Consortium, an organization overseeing development of industry-standard UALink accelerator-to-accelerator interconnections that supports up to 1,024 accelerators. While for now there is a limited choice of UALink switches, by 2029, there will be industry-standard switches offering different performance and capabilities, which makes the choice of NVLink as a scale-up fabric even stranger.

One possible explanation is that Apple is interested in considerably more than NVLink itself. NVLink Fusion is part of Nvidia's rack-scale and data center infrastructure architecture, which can combine NVLink scale-up connectivity with Nvidia's Spectrum-X Ethernet or Quantum-X InfiniBand scale-out networks, including switches equipped with co-packaged optics. Thus, Apple could potentially adopt Nvidia technology for both scale-up and scale-out connectivity instead of developing an entire data center networking stack of its own. This is merely speculation for now, but such an approach would effectively mean that Apple is building AI servers around significant portions of Nvidia's data center architecture while retaining its own processors and not using Nvidia accelerators. If this happens, this will be a testament that Nvidia is now setting de facto standards for AI data centers, no matter which AI accelerators and CPUs are used.

Burying the hatchet?

Without a doubt, Nvidia is a leading supplier of data center hardware, so it is logical for Apple to work with the company if the two companies are indeed working together on Apple's data center platform.

Apple and Nvidia are not exactly good partners. The feud between the two companies began in the early 2000s, when Steve Jobs accused Nvidia of infringing on Pixar's patents on which Nvidia responded that it owned more graphics IP than Pixar and therefore could sue the company. Later on, Apple and Nvidia had disagreements over GPU design decisions that the latter supplied to the former. However, then came 'Bumpgate' as Nvidia supplied Apple and other PC makers defective GPUs in 2007 – 2008, did not acknowledge the problem, and then resisted fully compensating Apple and other PC makers for their repair costs, which is when the relationship between the companies got especially dire. Apple continued to use Nvidia GPUs till 2014 or 2015, at which point it switched to AMD's Radeon, and then abandoned discrete third-party GPUs altogether.

More recently, Apple started to use Nvidia's hardware again. The latest Siri AI is primarily powered by Apple Foundation Models developed in collaboration with Google using Gemini technology. Server-side inference runs through Apple's Private Cloud Compute architecture, and many of the workloads are hosted on Nvidia Blackwell GPUs in Google Cloud. Yet, using Nvidia hardware in the cloud and adopting the company's technologies for your own platforms is a completely different thing.

Architectural enhancements and significantly higher clock speeds enable Apple's A20 Pro application processor (AP), used in the company's latest iPhones, to deliver not only a substantial generation-to-generation performance boost but also to outperform leading desktop CPUs from AMD and Intel by up to a whopping 32% in the single-thread Geekbench 7 benchmark, setting the record for the highest single-thread performance. While high-end PC CPUs still have more oomph for multi-threaded workloads, the tiny A20 Pro is still faster than mainstream laptop CPUs even when many threads are involved.

Fastest smartphone SoC

Apple's A20 Pro system-on-chip (SoC) delivers 4,006 points in single-thread and 11,460 points in the multi-thread Geekbench 7 benchmark, which represents a 23.3% higher ST performance and 27.1% higher MT performance compared to the immediate predecessor, the Apple A19 Pro, according to an early submission (which may or may not demonstrate performance of actual A20 Pro-based products, so take the results with a grain of salt).

The new SoC delivers the highest generation-over-generation performance improvement for Apple's smartphone processors in years and is currently the highest-performing mobile AP. Furthermore, the A20 Pro beats AMD’s 16-core Ryzen 9 9950X3D by 26% and Intel’s Core i9-14900KS by 32% in single-thread performance.

When it comes to the single-thread Geekbench 7 benchmark, Apple's A20 Pro outperforms its closest rivals — Qualcomm's Snapdragon 8 Elite Gen5 (SM8850) and Xiaomi's XRing O3 — by 31.5% - 33.7%. In fact, both SM8850 and XRing O3 perform on par with Apple's two-years-old A18 Pro. The six-core A20 Pro also beats the eight-core SM8850 by 12.2% in multi-thread workloads in Geekbench 7 and offers roughly similar multi-thread performance to the 10-core XRing O3.

Compared with other flagship smartphone processors, Apple's A20 Pro holds a commanding lead in Geekbench 7. It is 76% faster in single-thread and 48% faster in multi-thread performance than MediaTek's eight-core Dimensity 9400, while it nearly doubles the performance of Google's eight-core Tensor G5, with advantages of 99% and 96%, respectively. But the most striking gap of A20 Pro is with Huawei’s Kirin 9050 Pro: Apple's flagship is 290% faster in single-thread and 139% faster in multi-thread Geekbench 7 workloads.

A great laptop CPU

While Apple's A20 Pro continues to feature 'only' six cores like many generations before it, this time around the processor packs two 'super' desktop-class general-purpose cores running at up to 4.93 GHz, four efficiency cores running at lower clocks, and a memory interface that delivers +50% higher memory bandwidth compared to its predecessor (allegedly using a 96-bit memory I/O).

The architectural enhancements of advanced CPU cores running at nearly 5 GHz, along with a more capable memory subsystem, not only enable a massive generational performance uptick, but also allow the chip to offer unbeatable single-thread performance and massive multi-thread performance that is comparable to that of laptop CPUs, including previous-generation laptop CPUs from Apple.

Indeed, Apple's A20 Pro is 7% faster than M5, 20% faster than M4, and 43% faster than M3 in single-thread performance. Its six-core design cannot match its multi-thread performance, trailing the 10-core M5 by 39% and the 10-core M4 by 27%. Yet, it is only 5% behind the eight-core M3.

When compared to Intel's Panther Lake, the A20 Pro is 48.7% faster in single-thread performance than the flagship Core Ultra X9 388H, yet the 16-core Panther Lake processor is 63% faster in multi-thread workloads. Against lower-end Panther Lake parts, the A20 Pro is 74% – 88% faster in ST workloads and even leads the Core Ultra 5 325 and Ultra 5 332 by 3% and 64%, respectively, in multi-thread benchmarks.

The particularly striking results of Apple's A20 Pro are the 26% – 33% single-thread advantage over flagship AMD and Intel desktop CPUs, though the desktop processors remain dramatically faster in multi-thread workloads.

First 2nm smartphone SoC

When Apple transitioned to TSMC's N3B (3nm-class) process technology from N4 (4nm-class) with its A17 Pro SoC back in 2023, the new processor was barely 9.8% faster in ST and 6.8% faster than its predecessor A16 Bionic. By contrast, with its first 2nm smartphone SoC made on TSMC's N2 node, Apple offers a massive performance boost over the A19 Pro produced on N3P.

Indeed, Apple's A19 Pro packs two 'desktop-class' 'super cores' whose design is 'driven in part by increased front-end bandwidth, a new cache hierarchy, and enhanced branch prediction,' as Apple described its 'super cores' inside the M5 processor earlier this year. Such architectural enhancements obviously massively increase performance in single-thread workloads at the cost of increased die size, transistor count, and power. Apparently, N2 enabled Apple's designers to squeeze two desktop-grade CPU cores into a smartphone SoC.

Speaking of M5, it is noteworthy that A20 Pro delivers 7.1% higher single-thread performance than M5 while running at a clock speed that is 7.1% higher than that of M5, which is probably a good indicator that Apple's A19 Pro uses the same 'super cores' as M5.

While some may consider using PC-grade general-purpose CPU cores in a smartphone chip an overkill, Apple is known for using and supporting PC technologies in its mobile SoCs (NVMe, PCIe, DisplayPort-over-USB-C, hardware virtualization, etc.). Keeping in mind that Apple also uses A-series SoCs inside iPads and inexpensive laptops, it makes a great sense to have these technologies in its smartphone application processors. With desktop-grade cores inside the A20 Pro, the company greatly expands use cases of these CPUs while also solidifying their position in traditional segments that they will address in the coming quarters.

Without any doubts, Apple's transition to TSMC's N2 starts with a massive general-purpose performance increase, driven by 'fat' super cores and a memory subsystem featuring 50% more bandwidth compared to the A19 Pro. Over the next few weeks, we are also going to learn how Apple upgraded the GPU, NPU, and other aspects of the A20 Pro, and we are going to find out whether the upgrades are as impressive or incremental. In any case, so far, the A20 Pro looks very good.

We reported on the new Apple A20 Pro system-on-a-chip (SoC) for smartphones yesterday, an integral attraction within Apple’s first foldable, the iPhone Duo, and in the iPhone 18 Pro devices. Now the first A20 Geekbench 6 benchmark results are starting to pop up online, and they’re very impressive, particularly in single-core performance. If the result spotted by Longhorn is a typical one, the 4,719 single-core and 12,677 multi-core scores mean the new A20 is around 25% faster than its predecessor. Its single-core score can also make some of the best PC CPUs look anemic.

Apple’s official performance claims are interesting, as usual, but we’re always happy to see the third-party performance indicators start to emerge ahead of independent reviews. Heralding its new 2nm silicon yesterday, Apple might have actually understated the boost the A20 can deliver, with the official line about this “desktop-class” processor being the “fastest CPU in a smartphone,” and a claim that it is ‘just’ 20% faster than the previous gen. However, Geekbench isn’t the best indicator of real-world performance, and this is just a sample of one to sprinkle salt upon.

What are numbers without relevant comparisons, though? For more perspective on Apple’s newest silicon, which might also be thrown into a new Neo laptop (or desktop) in the coming months, check out the table below.

Above, we’ve pitted the Apple A19 Pro from last year’s iPhone 17 Pro as the second comparison column entrant. Apple has worked on multiple angles to deliver improvements over last year. It says that it has both new super-cores and efficiency cores in play. Then there’s the refined 2nm process and the faster clocks, too.

For some wider context, we’ve also tabulated one of Apple’s newest M5 computer chips, a modern Qualcomm Snapdragon Elite X2 laptop chip, and the AMD Ryzen 9 9950X3D2, for a fun desktop PC angle. Less fun is the comparison with the Asus Zephyrus G16 2024 laptop I’m using now, with an AMD Ryzen AI HX 370 chip. Its Geekbench 6 scores of roughly 2,800 / 14,500 are easily outclassed by Apple’s new smartphone processor in 1T tests, but retain a little dignity by winning by ~1,800 points in nT tests.

Apple is opening up pre-orders for its new iPhones with A20 silicon shortly, with retail release on Friday, September 18. It usually lifts review embargoes a few days before retail. We should therefore see a broader range of benchmarks and tests from good sources in the coming week.

A “beautifully preserved in working condition” Apple 1 computer motherboard signed by Steve Wozniak, accompanied by a letter signed by Steve Jobs, is up for auction next month at Bonhams in New York. The official estimate for this precious piece of computing history is US$500,000 - US$800,000, but we note that this ‘Hatfield’ Apple 1 previously sold for $686,000 in 2013. Let’s take a closer look at what you could get for your slightly under a million dollars.

This isn’t just an artifact to be encased in a glass cabinet. You could do that, of course, but Bonhams assures auction watchers that this Apple 1 is a working model. It also seems to come with all the essential accessories you'd need to get some computing done.

The motherboard is the centerpiece of the sale, of course. It is labeled “Apple Computer 1 / Palo Alto. Ca. Copyright 1976,” on the component side, with ‘NTI’ stamp underneath (second batch), and the board is complete with a white ceramic MOS Technology 6502 microprocessor, keyboard interface and connector, 3 original Sprague ‘Big Blue’ capacitors, 8KB RAM, expansion connector, cassette board connector, original Apple cassette interface board labeled ‘Apple 1 Cassette Interface Copyright 1976,’ explains the auctioneer description of the lot.

A period-appropriate monitor, cassette recorder, and pair of ASCII keyboards make up the 'Hatfield' Apple 1 hardware offering. Then there is also a selection of manuals, which would have come with the hardware bought in the late 1970s. An original Apple-1 Operation Manual; an original Preliminary Apple BASIC Users Manual with a green coversheet bearing Apple's original Newton logo, and an original Apple 1 Cassette Interface manual are included in the sale.

Another attraction of this auction is undoubtedly the Steven Jobs-signed letter. It is a letter from Jobs (dated 1978) to the original owner, Fred Hatfield, authorizing the exchange of the Apple 1 for an Apple II 4k board. The letter came in response to Hatfield complaining about the lack of software for the Apple 1. Jobs signed the letter making the offer, and it includes his title at the time: Vice President, Operations of Apple Computer Inc.

We’ve reported on Apple 1 auctions before, and this Hatfield sample is one of the most desirable Apple hardware artifacts, being from a batch personally designed and hand-built - then signed - by Steve Wozniak. The Apple 1 computer was first demonstrated in early 1976. Fellow Homebrew Computer Club members and Byte Shop visitors would buy up the first batch of (50) Apple 1 devices in the summer of 1976. Fred Hatfield, a computing enthusiast working at Computer Data Systems in Columbus, Ohio, bought his model sometime in 1976 from the second batch made available. In total, less than 200 Apple 1 systems were made.

The Hatfield Apple 1 last changed hands in 2013, achieving $686,000 at auction. Now, 13 years later, the estimated range of US$500,000 - US$800,000 might seem low, but sometimes auctioneers do this to catch broader interest. This Apple 1 will be sold as part of the History of Science, Technology & Computing auctions at Bonhams, New York, between October 4 and 14. You will be able to take part in the auction online.

The AI-apocalypse is upon us, and regular consumers are paying that price in a big way. Apple went on a price-hiking spree in late June, leaving only iPhones and AirPods untouched. The bulk of Apple’s computing devices saw price increases ranging from a few hundred bucks to $1,300. Apple is now ushering in a new wave of price increases with the refreshed Mac mini and Mac Studio.

The new M6 Mac mini is now priced at $899, a $100 increase over the already inflated $799 price for the M5 version on June 25th. To put those increases in perspective, the M4 Mac mini had a base price of $599 when it originally launched in October 2024.

The new $799 base price is an even more bitter pill to swallow, given that the headless computer still comes with only 256GB of base SSD storage. One would think that the $300 price increase over the course of two months would at least lead to a doubling of the base storage, but the ongoing memory/storage shortage makes that wish unthinkable; at least in Apple’s eyes.

If you want to see more eye-watering pricing, take a look at the hardware upgrade costs for the base Mac mini. The system comes standard with 16GB of unified memory, but an extra 8GB (taking the total to 24GB) costs a whopping $200. If you want 32GB (the maximum allotment for the M6 chip), that’ll cost you an extra $400.

Things don’t look much better on the storage side, with an upgrade from 256GB to 512GB costing $200. The 1TB and 2TB storage options cost $500 and $1,000, respectively. Apple’s value-leader Mac mini is now in uncharted territory. A maxed-out Mac mini (M6, 32GB, 2TB) costs $2,229.

Even the M5 Pro version of the Mac mini couldn’t escape the AI tax, as the base configuration now costs $1,699, up from $1,599 for the M4 Pro version (June 25 pricing) and a far cry from the $1,399 price of the previous generation earlier in the year. That price gets you a 15-core GPU and a 16-core GPU, while stepping up to the uprated 18-core CPU/20-core GPU version adds $200. Memory upgrades are deflating to the psyche, as upgrading from 24GB to 48GB adds $600, while the 64GB option is $1,000.

Storage options are absolutely mind-boggling: a 1TB SSD requires an additional $300 outlay, while the max 8TB SSD configuration is listed at $3,800. The range-topping configuration (M5 Pro, 18-core CPU/20-core GPU, 64GB, 8TB) costs $6,699.

If there’s any consolation, the new Mac Studio holds the line on pricing ($2,499) in its base configuration (18-core CPU/32-core GPU, 36GB, 512GB). The price stasis is somewhat surprising given the hardware upgrades, including a new M5 Max SoC and PCIe Gen 6 SSD. However, upgrades are pricey, with the 18-core GPU/40-core GPU variant of the M5 Max adding $300 to your bill. The upgraded SoC is the only way to unlock additional memory options (48GB, 64GB, 128GB).

Stepping up the Mac Studio with M5 Ultra more than doubles the base, entry price of the M5 Max config at $5,499. This gets you a 30-core CPU and 64-core GPU, while the upgraded 36-core CPU/80-core GPU is a $1,300 option. 96GB of RAM comes standard, while 256GB is a $4,000 upcharge. SSD options are equally expensive: 2TB costs $500, and 16TB costs $7,500.

No matter which way you slice it, those firmly entrenched in the Apple hardware ecosystem are paying a heavy price for memory/storage shortages. Apple customers are doubly affected, as the company’s traditionally outrageously priced memory and storage upgrades have become even more expensive, putting Apple’s previously value-driven models like the Mac mini out of reach of bargain shoppers.

Apple has today unveiled its next-generation M6 Apple silicon chip, as well as a powerful M5 Ultra. The new chips will debut in new versions of its Mac Mini and Mac Studio, respectively, available to pre-order from today.

The M6 is Apple's first chip on a 2-nanometer process, utilizing a 12-core CPU with two super cores, four performance cores, and six efficiency cores. The chip also features a 12-core GPU and a 12-core GPU featuring neural accelerators, alongside a "Dual 16-core Neural Engine." The top version of the chip has 170GB/s of memory bandwidth.

That dual 16-core neural engine means that there are indeed two neural engines. There's a connection between the engines, allowing them to run one model quickly across both, or two separate models independently.

The M5 Ultra is Apple's first chip using its UltraFusion technology to form a quad-die architecture (The M5 Max chips it is connecting both used two dies). The chip goes up to 36 cores on the CPU and up to 80 cores on the GPU, with 1.2TB/S of unified memory bandwidth, which Apple says is 50% higher than the M3 Ultra. Apple also claims the 36-core CPU with 12 super cores and 25 performance cores will offer "up to 1.25x higher single-threaded performance and up to 1.3x higher multithreaded performance than M3 Ultra."

The M5 Ultra has a 32-core neural engine, but this is different from the dual 16-core neural engine on the M6. Because the M5 Ultra uses a pair of M5 Max chips over UltraFusion, the 32 cores across the dies to run two models in parallel for increased performance.

Both M6 and M5 Ultra feature GPUs with neural accelerators and their fastest cores, marking a leap forward for Apple's graphics processors when it comes to AI.

Apple users who have been maxxing out previous versions of the Mac Mini will be capped at 32GB of memory on the Mac Mini with M6, like a choice the company made due to the component shortage that has been affecting the entire industry. If you want 64GB like the M4 Pro, you'll need the M5 Pro model. But the M5 Ultra in the Mac Studio won't have that problem, with Apple offering up to 512GB of unified memory on that chip.

M6 also features updates to hardware-accelerated ray tracing, dynamic caching, and the shader core, which the company claims will allow for faster rendering and higher frame rates in games.

Mac Mini and Mac Studio

The new Mac Mini and Mac Studio aren't seeing any design changes, though there are substantial internal changes. Both systems are available for pre-order today, but won't arrive to customers and in retail stores until September 22.

Besides the M6 version of the Mac Mini, there will also be versions with M5 Pro. Both models will be bumped to Wi-FI 7 and Bluetooth 6, as well as 2.5Gb Ethernet (up from 1Gb), with a 10Gb option available as an upgrade. The Mac Mini also supports genlock via USB-C, synchronizing displays and cameras to prevent video drift during broadcasts.

For those working with AI, Thunderbolt 5 on the M5 Pro Mac Mini will allow customers to run clusters for on-device models.

The Mac Mini with M6 will start at $899, while the M5 Pro option will begin at $1,699 (each drops $100 for education pricing). Given that the Mac Mini with M4 started at $599, it's seeing substantial price hikes here.

The Mac Studio will come with M5 Max or the new M5 Ultra chip, and Apple is positioning it for on-device AI. It will now offer up to six Thunderbolt 5 ports with support for as many as eight displays. Like the Mac Mini, it's also getting Wi-Fi 7 and Bluetooth 6 with Apple's N1 chip, which debuted in the iPhone 17 lineup. Like the Mac Mini, the Mac Studio also supports genlock for perfectly synced video.

The Mac Studio with M5 Max will start at $2,499 ($2,299 with an education discount), and the M4 Ultra model will start at an eye-watering $5,499 ($5,099 with an education discount).

Both systems will be compatible with macOS 27 Golden Gate, including Siri AI and refinements to the Liquid Glass design introduced in macOS 26 Tahoe.

The Dutch National Cyber Security Centre (NCSC-NL) reported on August 12 that attackers are actively exploiting CVE-2026-65400, an authentication bypass in macOS Screen Sharing, to compromise Macs with port 5900 exposed to the Internet. In every case reported to the agency, attackers obtained root access and installed a Monero cryptocurrency miner. Apple patched the flaw on August 6 in an out-of-band update covering macOS Tahoe 26.6.1, Sequoia 15.7.9, and Sonoma 14.8.9, but the bug's official severity has since been rewritten, with CISA raising its CVSS score from 7.1 to 9.8 critical on August 14, now assessing the attack as automatable.

NCSC-NL first flagged the vulnerability in an advisory on August 7, a day after Apple's patch, urging organizations to update immediately. The August 12 revision added that public proof-of-concept code is now available and that active abuse had been observed on multiple internet-exposed systems.

Technical details of the bug were presented at last week's Black Hat conference, according to Ars Technica, alongside a video of the exploit in action. The root-level access attackers gained in the reported incidents matches the level of control researchers demonstrated in May, when they bypassed Memory Integrity Enforcement on Apple's M5 silicon with AI assistance.

NVD's change log for CVE-2026-65400 shows CISA initially scored the bug at 7.1 on August 6, using a vector that assumed an attacker needed low-level privileges and could achieve only partial impact. On August 14, the agency replaced that vector with one requiring no privileges and granting full compromise of confidentiality, integrity, and availability, raising the score to 9.8. A day later, CISA's decision record for the flaw flipped from "not automatable" to automatable, an assessment consistent with unattended Macs being rooted at scale for coin mining.

The flaw still isn't in CISA's Known Exploited Vulnerabilities catalog as of this writing. The same decision record also still lists exploitation as "none," despite the NCSC-NL report. Users who can't update immediately can turn Screen Sharing off under System Settings > General > Sharing.

Apple's advisory says an attacker on the network may be able to "authenticate to Screen Sharing without valid credentials," and describes the fix as improved state management during authentication. Screen Sharing is the VNC-based remote desktop service built into macOS, listening on TCP port 5900, which is disabled by default. The August 6 update comes just 10 days after Apple’s July 27 security round and fixes only this single CVE, representing the second Screen Sharing patch in a month. CVE-2026-43760, which required valid credentials to exploit, was fixed in the late-July releases.

Apple used to be among a few companies willing to buy memory and other components worth billions of dollars under long-term supply contracts at fixed prices. But the artificial intelligence era represents a new reality with new purchasing champions, marking a tectonic shift in the high-tech world. Alphabet, Microsoft, Meta, and Amazon have purchase commitments totaling about $2 trillion, and a significant portion of these commitments are for memory, according to estimates by analyst Claus Aasholm. While the commitments are approximate, span many years, and should be generally taken with a grain of salt, they still reflect the direction the industry is moving.

Combined purchasing commitments from the four major hyperscalers shown in the chart —Amazon, Alphabet, Meta, and Microsoft — reached nearly $2 trillion by Q2 2026, with Alphabet and Microsoft accounting for the overwhelming majority of the total.

The rapid expansion suggests several major findings. Firstly, the AI infrastructure race is accelerating, not stabilizing. Secondly, AI infrastructure investments are driven by a handful of hyperscale cloud service providers (CSPs) whose long-term procurement commitments now vastly exceed those of traditional consumer electronics companies such as Apple.

Thirdly, memory has become a strategic asset — perhaps a competition weapon — rather than a commodity. Fourthly, suppliers of memory — both 3D NAND and DRAM — are gaining pricing power. Finally, demand for memory will likely drive major capacity expansion at Micron, Samsung, and SK hynix, even though so far these companies have been exceptionally disciplined about their capacity investments.

Almost $2 trillion commitments

Google shows by far the most aggressive increase in purchasing commitments, rising from roughly $140 – $150 billion in Q3 2025 to around $811 billion by Q2 2026 (though these are total purchase commitments by Alphabet, not specifically memory purchase commitments), while Microsoft follows a similar trajectory and reaches approximately $678 billion in total obligations, which includes, but is not limited to memory.

Meta is also ramping commitments substantially to around $349.3 billion (again, these are total commitments), whereas Amazon increased its commitments more gradually to roughly $130 billion. By contrast, Apple — which makes the world's most popular smartphone, and which was the largest consumer of memory just a couple of years ago — remains almost flat throughout the period at approximately $57 billion (of which $56.2 billion is payable within 12 months). Apple's commitments fall well short of Nvidia's commitments of $119 billion.

Again, we are talking about total purchase commitments, which include foundry capacity, 3D NAND, and DRAM memory, but are not limited to them. Alphabet, Amazon, Meta, and Microsoft all build custom silicon and custom servers, so a significant portion of these commitments is to various EMS providers.

While $1.968 trillion of purchase commitments for memory and storage alone would be an absurdly large amount of money, a huge portion of these commitments consists of contract manufacturing obligations as well as memory chips. This suggests that the foundry, 3D NAND, and DRAM markets are entering a new phase in which hyperscalers are willing to make vastly larger forward purchasing commitments than traditional consumer-electronics companies, giving suppliers a strong incentive to prioritize customers prepared to secure future capacity on that scale.

Strategic assets

While Claus Aasholm's chart is explicitly dedicated to memory, it does describe total purchase commitments of tech giants, so the chart can reasonably be read as evidence that memory and capacity at TSMC, Samsung Foundry, and GlobalFoundries are becoming a strategic asset rather than merely another component to procure at the best available price.

AI infrastructure requires enormous quantities of AI accelerators, DRAM (including HBM), and 3D NAND. Meanwhile, the supply of high-end memory (HBM) is constrained by fab capacity at major DRAM makers, whereas the supply of AI accelerators is constrained by both wafer capacity and foundries and packaging capacity at foundries and their OSAT partners. As a result, hyperscaler CSPs have an incentive to lock in supply years ahead, even if doing so requires exceptionally large purchasing commitments.

That also changes the relationship between semiconductor suppliers and their customers. In theory, a company willing to guarantee hundreds of billions of dollars of future purchases can effectively help underwrite expansions of foundry, memory, and advanced packaging capacity and, in return, secure priority access to scarce products and future process technologies. In reality, TSMC can well afford capacity expansion using the money it gets from hyperscalers and give priority to its largest customers. In this environment, access to DDR5, HBM, and 3D NAND memory becomes part of the competitive advantage rather than merely a procurement exercise.

This is also what makes Apple's position in the graph interesting: its purchasing commitments barely move while those of Alphabet, Amazon, Meta, and Microsoft surge. If the trend continues, Apple may remain one of the world's largest semiconductor buyers in absolute terms, but the question is whether it will be among the key customers that foundries, memory makers, and OSATs plan their future capacity expansions.

An inflection point

Perhaps the most interesting takeaway of the findings revealed by long-term purchase commitments is that the industry's center of gravity appears to have shifted.

During the smartphone era, foundries (well, TSMC has won) and memory suppliers often competed aggressively for Apple's business because of its enormous purchasing power. Today, hyperscalers building AI infrastructure are making purchasing commitments that dwarf those of traditional CE companies like Apple, which may well represent a strategic inflection point akin to the one Andy Grove described in his 'Only the Paranoid Survive' book.

Will this tectonic shift result in prioritization of customers capable of enabling future capacity expansions through massive long-term purchase agreements, or will foundries and memory makers remain more or less disciplined with their capacity expansions so as not to lose a lot when demand declines? This is a question that has yet to be asked.

In any case, the AI megatrend has transformed semiconductors — from foundries to advanced packaging and from DDR5 to HBM4 — into strategic assets that can no longer be treated as ordinary components procured on demand. And this is something that will continue in the long run.

20 years ago today, Steve Jobs took to the Apple WWDC stage to unveil the first-ever Mac Pro desktop computer. As well as marking the beginning of a whole new Mac line, this was the first Apple workstation desktop to transition from PowerPC to Intel x86. Apple had already introduced Intel-powered versions of its iMac, Mac Mini, and MacBook at the start of 2006; now it was replacing the mighty PowerMac G5 tower.

While readers might remember most clearly, and even fondly, Intel-era Mac Pro computers like the ‘trash can’ or the ‘cheese grater,’ the first Mac Pro with an x86 chip looked almost the same as its G5 ancestor.

Of course, the Apple WWDC presentation was all about the new processors, and the new machine was crowned the performance king, claimed to “deliver up to twice the performance of the Power Mac G5 Quad.” The top configuration would come with two new dual-core Intel Xeon processors running up to 3.0 GHz.

“Apple has successfully completed the transition to using Intel processors in just seven months — 210 days to be exact,” boasted Apple CEO Steve Jobs. “And what better product to complete it with than the new Mac Pro, the workstation Mac users have been dreaming about.”

Key specs of the first Mac Pro were:

• Two Dual-Core Intel Xeon processors running up to 3.0 GHz, each with 4MB of shared L2 cache and independent 1.33 GHz front-side buses
• Nvidia GeForce 7300 GT with 256MB GDDR3, configurable up to ATI Radeon X1900 XT and the Nvidia Quadro FX 4500, both with 512MB of video memory
• 1GB of DDR2-667 fully buffered memory, expandable to 16GB
• 250GB HDD, expandable to 2TB using the four bays
• 16x SuperDrive with double-layer support (DVD±R DL/DVD±RW/CD-RW), with an extra bay for two ODDs
• 2x Gigabit Ethernet, Optional Wi-Fi 4, Optional 56k V.92 USB modem, Optional Bluetooth 2.0+EDR
• 5x USB 2.0 ports, 2x FireWire 800 ports, 2x FireWire 400 ports, various audio I/O
• 4× PCIe 1.0 slots, one double–wide for the best graphics cards of the time
• Compatible with Mac OS X Tiger at launch, supported to 10.6 Snow Leopard

Three processor choices were available at release in early 2007. As well as the top-end model with a pair of dual-core Core Xeon 5160 CPUs running at 3.0 GHz, there was a Core Xeon 5150 x2 (2.66 GHz), and a Core Xeon 5130 x2 (2.0 GHz) model. If an off-the-peg model wasn't quite right, Apple proudly claimed the online configurator allowed for "more than 4.9 million possible configurations."

To give you some pricing context, Apple released the Mac Pro with a Core Xeon 5150 x2 (2.66 GHz) and base specs bullet-pointed above for $2,499, including a Mighty Mouse and Apple Keyboard.

Apple’s last Mac Pro was a 2023 model with the M2 Ultra chip. Then, in March this year, Apple confirmed there would be no future revisions of this desktop workstation flagship. The line had been effectively replaced by the Mac Studio, as a key reason for the Mac Pro to exist (upgradeability via PCIe expansion) had been sidelined by Thunderbolt 5.

Apple reportedly has around $1 billion in "processor wafers on the shelves awaiting packaging" amid ongoing memory shortages, putting pressure on the firm as it nears the launch of the iPhone 18. Tim Culpan of Culpium says Apple is "scrambling" for memory supply ahead of release, working with suppliers to get memory ready so it can finish packaging the A20 Pro and C2 chips that are reportedly inside the iPhone 18 lineup.

It's an interesting peek behind the curtain of the current supply chain limitations, which reach beyond simply DRAM chips. With fully-integrated SoCs across both desktop and mobile, Apple is uniquely vulnerable to supply chain limitations, not only in DRAM supply, but also in packaging capacity to bring together DRAM chips with processor wafers.

According to Culpan, Apple primarily sources DRAM from Micron, with smaller engagements from SK Hynix and Samsung. Micron (along with the other major memory players) has reportedly sold capacity for DRAM and HBM through 2027. A flurry of long-term agreements (LTAs) have cropped up over the past several months, guaranteeing supply to certain companies for several years.

The A20 Pro is the first chip to use TSMC's N2 process, and it has been progressing well, with TSMC reportedly sitting on around $1 billion worth of Apple wafers awaiting packaging with DRAM. Packaging happens with TSMC's new Wafer-Level Multi-Chip Module (WMCM) process, according to Culpan, which is similar to the CoWoS packaging but for mobile parts.

Culpan reports that "Apple and its assemblers remain confident that they can meet initial demand," but DRAM shortages could put pressure on shipments past the initial launch. Apple reportedly plans to ship around 200 million iPhone 18 units across the lineup throughout the device's lifespan. The iPhone 17 range saw record shipments last year, surpassing 245 million devices shipped in 2025.

Apple is one of several companies reportedly looking to alternative suppliers for DRAM, namely China's CXMT. The Chinese memory supplier has become a hot topic in the halls of the U.S. government over the past several weeks, with Apple reportedly lobbying for clearance as the government weighs the option to ban CXMT.

Apple and OpenAI are heading to court in a clash of two of the world's largest and most well-funded companies. With teams of lawyers and bottomless pockets to fund them, these two industry titans are squaring off over claims of theft of trade secrets. Apple is asking the courts to block its former employees from revealing information to their new employer and limit OpenAI's development of an AI device based on Apple technology. Unsurprisingly, OpenAI denies any wrongdoing and is asking the judge not to disrupt its business activities.

"Apple's getting this wrong" is the name of the blog post OpenAI posted in response to Apple's lawsuit. Although it begins by calling Apple one of the greatest companies of all time, it goes on to call the legal action "sad," suggests Apple mixed up Asian employee names, and doesn't want its trade secrets anyway.

OpenAI went on to publish text message chains and email correspondence, suggesting it's keen to be tried in the court of public opinion more than any other setting.

Hey, that's mine!

Apple alleges that as many as 13 of its former employees may have passed along sensitive trade secrets to OpenAI, including sharing details of unannounced products and sharing screenshots of confidential Apple documents. This is an expansion of its original claims from July, when it alleged two former employees may have shared confidential information with OpenAI. It has now filed a concurrent motion seeking expedited discovery, demanded that two former employees, Chang Liu and Tang Yew Tan, sit for depositions, as well as key OpenAI employees and representatives of OpenAI's hardware division, io Products.

Apple is also asking for a preliminary injunction, which would bar former employees from accessing, acquiring, or disclosing any confidential information about the company as the case proceeds.

Considering a case of this magnitude, between two companies with effectively infinite money to spend protecting themselves, that could take months or years to resolve, that preliminary injunction is key and could mean the lawsuit does enough damage to be effective for Apple's goals.

Regardless, OpenAI is in full revolt over the process and has posted not only a damning denial of all claims, but some evidence it feels exonerates its position, too.

Getting ahead of the competition

The cynical argument for this lawsuit is that it means the two companies are battling it out in a courtroom instead of in the consumer retail space. Apple is first and foremost a consumer device company, and to date, OpenAI hasn't competed there. But it's planning to, and its various hires from Apple and elsewhere, including buying up Jony Ive's hardware company, suggest it's going to release something with the kind of smart consumer focus that is typically Apple's bread and butter.

Apple might have been a little late to the AI game, but that's played to its advantage. But if OpenAI were to use its strong mindshare, deep pockets, and advanced AI model design and inferencing compute base to offer something targeting mainstream consumers, it's stepping on Apple's toes. This lawsuit could be a way to stop that in its tracks.

That's not to say the lawsuit may not have merit. The entire AI industry has used aggressively expensive hiring practices to try to get ahead. Meta spent $15 billion on getting Alexandr Wang and ScaleAI's data, and AI companies are reportedly hounding promising students years ahead of graduation to lock them into future contracts.

If you're building a smart consumer device, though? Grabbing ex- or even current Apple employees wouldn't be a bad way to do it. Especially if they brought along information on unannounced Apple products that may or may not be related to an AI consumer device.

First to market matters, too. Although Apple hasn't shown any AI-first standalone devices yet, if it were planning to, slowing OpenAI down with these injunctions might be enough to let it catch up or even get ahead.

Clock's ticking

For now, the lawsuit is very much he-said, she-said. Both companies have enough money and resources to make this fight messy, but what they don't necessarily have is time. Especially OpenAI. It's on the hook for hundreds of billions in compute investment, has just cut the price of some of its models, and has yet to post anything close to a profitable quarter.

Apple isn't quite so constrained, but is also seeing its traditional consumer base eaten by rapidly appreciating electronics costs. Apple already sells its products at a premium. But what can it do if prices for components push its products out of the luxury affordable into the truly unaffordable?

That may not happen, but it's a future it has to consider as AI, and the money spent on its promised future continues to have an outsized impact on global economies. The outcome of this lawsuit is likely to leave a mark, too.

Pick up an Apple MacBook Pro 14 with M4 Pro chip and a hefty 48GB of RAM for $2,999, thanks to a nice $900 discount at Best Buy. This top-spec MacBook Pro may feature the older M4 Pro chip, but it still packs hefty memory, storage, and performance specs. If you need a laptop with ample RAM for your productivity applications, this deal could be of interest to you.

● Check out this deal at Best Buy

This 14-inch MacBook Pro uses the Apple M4 Pro chip with 14 cores for CPU tasks and 20 cores for GPU operations. This model comes stacked with 48GB of unified memory and a 2TB SSD. The screen on the MacBook Pro features a Liquid Retina XDR nano-texture display with up to 1600 nits peak brightness, one of the brightest and best-looking screens we had reviewed at the time of testing, helping to earn this M4 Pro laptop an Editor's Choice award in our review.

We tested the 14-inch MacBook Pro M4 in November, 2024. It was the first laptop we tested with Apple's M4 chip. The SKU of the laptop we tested was less powerful than the model we are featuring today, featuring only 16GB of RAM and 1TB of SSD storage compared to the 48GB and 2TB in this model.

Apple's M4 chip showed strong performance on benchmarks, easily competing with the competition in both single and multicore workloads. On Geekbench, the new Mac achieved a single-core score of 3,807 and a multi-core score of 15,114.

This particular model of the MacBook Pro 14 with M4 Pro does lean on the expensive side, even when discounted to $2,999, but that's because it is stacked with a massive amount of RAM and a larger SSD, which Apple does charge a premium for. As this is unified memory and soldered to the mobo, there is no upgrade path later down the line, so if you want a laptop with a larger amount of memory, then you need to choose the amount you want at the point of purchase.

Apple filed a lawsuit against OpenAI in July, alleging that the latter stole trade secrets through former Apple employees. The startup has finally publicly responded to the lawsuit today, posting a blog post on its website titled, “Apple is getting this wrong.”

“Apple is one of the greatest companies of all time, and built a reputation for obsessing over the smallest details. This careless, aggressive and oddly personal lawsuit sadly doesn’t live up to that reputation,” the company said in its blog post. It also added, “Apple’s request for a preliminary injunction is both based on false information and completely unnecessary because we do not have, nor want, any of their trade secrets.”

The case names OpenAI technical staff Chang Liu, chief hardware officer Tang Tan, OpenAI, and io Products as defendants. Liu worked as a senior electrical engineer at Apple, while Tan was its former vice president of product design for the iPhone, AirPods, and the Apple Watch. Interestingly, io Products, which was acquired by OpenAI last year, was founded by Tan alongside former Apple Chief Design Officer Jony Ive and other former Apple heavyweights.

The tech giant requested a preliminary injunction to stop the accused from accessing, acquiring, using, or disclosing the alleged stolen trade secrets. Cupertino said in its filing, “Apple will be irreparably harmed absent a preliminary injunction.”

OpenAI publicly responded to this legal request through its blog post, pointing out that many of its allegations were actually based on miscommunications and false information, and that this move is completely unnecessary. But whether what the AI tech giant claims is true or not, its research and operations on consumer hardware would likely be impaired if the judge grants Apple’s request.

Although OpenAI’s primary product is ChatGPT and its other LLM-powered services, it’s been rumored that it’s also working on AI-powered hardware. Its first product is expected to be some sort of portable smart speaker with built-in cameras and sensors. Although this isn’t an iPhone replacement, like the botched Humane AI Pin, some experts say this move threatens Apple’s standing as the biggest consumer electronics company on the planet. This is especially true as it still lags behind other tech giants when it comes to artificial intelligence, even though AI enthusiasts love using Apple silicon with its Unified Memory architecture.

Apple’s preliminary injunction is just the opening stages of what might be an epic showdown between two tech giants. It will likely take months (or even years) before this case reaches a judgment or a settlement, meaning a preliminary injunction will definitely have a massive impact on OpenAI’s hardware dreams.

Apple will pay even more for memory in the September quarter than it did in the June quarter, CEO Tim Cook told analysts on the company's earnings call on July 30, after memory costs accounted for more than the entire sequential decline in Apple's adjusted gross margin. Cook, speaking from Cupertino on his final call before John Ternus takes over as CEO, called the market "a hundred-year flood on the memory pricing." Apple's consolidated financial statements for the quarter ended June 27 put inventories at $11.09 billion, up 87% year over year, with $5.46 billion of cash consumed building that position over nine months. CFO Kevan Parekh said the benefit Apple gets from that carry-in inventory shrinks after September.

Apple held $5.93 billion of inventory at the end of June 2025 and $5.72 billion at its September fiscal year end, so the $11.09 billion reported for June 27 is a 94% increase in nine months. The cash flow statement puts the same movement at $5.46 billion consumed by inventories over those nine months, against a $1.22 billion release in the year-ago period. Measured against quarterly cost of sales of $54.65 billion, the June position works out to roughly 18.5 days of inventory, up from about 10.7 days a year earlier. Apple has run one of the leanest working capital positions in consumer electronics for two decades, but that has now been flipped on its head.

Parekh told analysts the carry-in inventory partially offset memory costs in the June quarter and will do so again in September, with a decreasing benefit beyond that. Cook said market pricing for memory keeps rising past September and that the effect on Apple's business could grow. Erik Woodring of Morgan Stanley asked whether Apple intends to pursue multi-year long-term agreements with suppliers at pre-agreed prices. Cook answered the second half of the question, on pricing philosophy, and left the agreements question alone.

Company gross margin was 49.3% in the March quarter and 48.1% in June, once the roughly two percentage points of tariff refunds come out, Parekh said, and more than 100% of that 120 basis point decline is explained by memory costs. Apple guided September gross margin to 47% to 48%, including about one percentage point of tariff refunds, putting the adjusted midpoint near 46.5%, and Parekh attributed that step down to the same dynamic. Foreign exchange was a factor in the June quarter but not the main one.

The cost of sales for products rose 8.1% year over year to $47.15 billion, while product revenue increased 18.1% to $78.68 billion, which lifted product gross margin to 40.1% from 34.5%. The Mac and iPad price increases that came in June, the tariff refunds, and the stockpile together more than covered the memory increase in the reported quarter. On the other side of the trade, SK hynix ran a 76% operating margin over the same three months on revenue of 79.32 trillion won and operating profit of 60.54 trillion won, the company said recently, and guided third-quarter DRAM bit shipments up only about 10% sequentially. TrendForce expects conventional DRAM contract prices to rise a further 13% to 18% in the third quarter.

Cook, asked whether Apple's push for sourcing flexibility is about securing volume or protecting its price points, said the DRAM market has three suppliers and that more of them would help on the supply side and perhaps on pricing, then corrected himself to say the pricing effect is unclear. He said Apple is "evaluating all options." The fourth supplier available to him is CXMT, which the Financial Times reported in July that Apple has begun testing DRAM for devices sold in China while lobbying Washington for clearance to use its parts more broadly. Representatives John Moolenaar and George Whitesides wrote to Commerce Secretary Howard Lutnick earlier this month, asking for purchases from CXMT and YMTC to be barred outright, including through allied supply chains. CXMT listed on Shanghai's STAR Market in July.

The unit supply constraints Apple flagged for the September quarter are a separate problem from memory pricing. Cook attributed them to the availability of the advanced process nodes Apple's SoCs are built on, after iPhone and Mac demand ran ahead of the company's plan, and said they'll hit iPhone, Mac, and iPad in September, against Mac primarily in the June quarter. Apple guided September revenue growth of 9% to 11% year over year, down from 16% in the previous quarter.

Our local AI explorations in 2026 have so far focused on two main platforms: Nvidia’s GB10, as seen in the DGX Spark and Dell Pro Max with GB10, and AMD’s Ryzen AI Max+ 395, aka Strix Halo, as seen in the Corsair AI Workstation 300 and the Ryzen AI Halo. We’ve generally favored GB10 systems for these kinds of local AI development sandboxes. Their solid all-around performance and broad AI software compatibility make them easy to love, even if raw LLM inference throughput isn’t that high.

But Apple’s Mac Studio is another compelling option for local AI trailblazers who want systems with large unified memory pools, powerful GPUs, and high memory bandwidth to enable faster tokens-per-second throughput than either GB10 or Strix Halo can provide. And until someone builds another unified memory SoC with a memory bus as wide as what Apple uses for its Max and Ultra chips, Apple Silicon is, in fact, the only game in town for more memory bandwidth from a chip of this design.

Why focus so much on memory bandwidth for local LLM performance to begin with? To keep it very short, the decode phase of LLM inference is virtually always sequential. Every output token (starting with the first one generated as part of the prefill process) has to be run through every layer in a model in order to generate the next token, and so all the active model weights in a layer have to be streamed to the GPU from its memory as the token makes its way through each one.

Because the amount of computation required for each individual token at each layer is tiny, the speed of the entire decode process basically becomes dependent on how fast those model weights can be streamed in from GPU memory. That makes systems with large memory pools (for holding large models) and high memory bandwidth (for streaming those model weights to the GPU) appealing, and Apple Silicon is the only platform that lets you have both characteristics at a (relatively) reasonable cost.

Apple’s last Mac Studio refresh was in March 2025 and marked the introduction of two versions of the system. The higher-end version uses the M3 Ultra SoC with 819GB/s of memory bandwidth across both the 28-CPU-core, 60-GPU-core mode and the 32-CPU-core, 80-GPU-core model.

The M4 Max version of the Mac Studio encompasses two versions with greater divergence in specs than just core counts. The base M4 Max includes a 14-core CPU (10 performance cores and four efficiency cores) with 410GB/s of memory bandwidth, and it can be upgraded to a 16-core CPU (12 performance cores and four efficiency cores) with 546GB/s of memory bandwidth.

We reviewed both of these systems at launch, but Apple generously lent us another M4 Max with the 16-core CPU, 40-core GPU, and 128GB of memory for LLM-specific tests. At the time, the price of that system was $3,699 or so, close to the DGX Spark’s $3,999 list price.

But as the RAMpocalypse has rolled on, the Mac Studio has become more limited in its configuration options and harder to find in stock. The M4 Max version we tested now tops out at 64GB of RAM, and lead times for that system are over two months out. Regardless, it’s the most comparable system that Apple has for the DGX Spark and GB10 platform more generally, at least back when dollar-for-dollar meant something in this market.

Before we get into performance results, we wanted to dig into the nitty-gritty of the M4 Max’s 40-core GPU, but Apple doesn’t share much (if any) detailed architectural information on its GPUs, including the number of shader ALUs that make up each GPU core, so it’s hard to directly compare the M4 Max to GB10 or the Radeon 8060S in the Ryzen AI Max+ 395.

The Radeon 8060S on board Strix Halo has 40 RDNA 3.5 compute units, each with 64 shader ALUs for a total of 2,560. AMD clocks the 8060S high, at 2900MHz, to achieve its roughly 14.8 TFLOPS of peak FP32 vector math.

Meanwhile, GB10 has 48 Blackwell SMs with 128 shader ALUs each for 6,144 total CUDA cores, and it too clocks relatively high, at 2450MHz in our testing. It offers a whopping theoretical 30 TFLOPS of FP32 vector compute.

Way back in the day when Apple first revealed the M1 GPU, however, the company disclosed that each core encompassed 128 execution units. Assuming that the same organization has carried through to current Apple GPUs, the 40-core M4 Max GPU would have 5,120 execution units or shader ALUs. That provides an idea of the scale of the GPU, at least, though without knowing clock speeds, we’re hesitant to speculate further about peak FLOPS or other theoretical performance measures.

Apple also only provides raw memory bandwidth figures for the M4 Max, but we can do some back-of-the-napkin math to surmise that the chip likely uses LPDDR5X-8533 on a 512-bit bus to achieve its 546 GB/s of rated memory bandwidth, twice as wide as both GB10 and Strix Halo. The lower-end M4 Max likely uses LPDDR5X-6400. In turn, the M3 Ultra likely uses a 768-bit-wide bus with LPDDR5X-8533 across both of its variants.

We covered the design of the Mac Studio in detail in our initial review, but it’s worth a quick tour to re-familiarize yourself with it. The top and sides of the system are largely featureless, save for an inlaid black Apple logo up top. The front panel has two 10Gbps USB Type-C ports and a UHS-II SD card reader.

The Studio’s thermal design uses a clever intake system that draws in air from vents around the system’s circular foot before exhausting it through a vent at the back of the chassis.

The M4 Max Studio we tested has a whopping four Thunderbolt 5 ports on its back panel with support for data rates up to 120Gbps. Although we’re not testing it today, you can enable RDMA over Thunderbolt 5 for low-latency clustering of Mac Studios using utilities like Exo. These ports also support Thunderbolt displays or DisplayPort 2.1 Alt Mode.

This system also has two USB 3 ports running at 5Gbps, 10Gb Ethernet, an HDMI 2.1 port, and a headphone jack, so it’s got all the connectivity you could practically want out of a compact desktop in 2026.

With those basics out of the way, let’s get to testing.

To test LLM inference performance, we once again turn to llama.cpp, which includes native Metal/MLX support on Apple Silicon when built from source. We used the llama-benchy project as our benchmarking harness.

We tested three models of various memory footprints and architectures: Qwen 3.6-35B-A3B, a recent mixture-of-experts model; Gemma 4 12B, a small dense model, and OpenAI’s gpt-oss-120b, a large mixture-of-experts model. We used Unsloth’s four-bit quantizations of these models for our testing.

Out of the gate with Qwen 3.6-35B-A3B, the Mac Studio delivers prompt processing speeds that beat Strix Halo but are still behind GB10, especially with longer contexts.

Flip over to the throughput results, and the M4 Max pulls ahead of both GB10 and Strix Halo, turning in higher tokens-per-second throughput than both those systems at all context depths.

But despite having double the memory bandwidth of GB10 and Strix Halo on paper, the real-world improvement in token throughput on the M4 Max is just 25% on Qwen 3.6-35B-A3B. That’s certainly welcome, but it’s not nearly the speedup you might expect with twice the memory bandwidth on tap.

The dense Gemma 4 12B, on the other hand, really seems to let the Mac Studio’s bandwidth advantage shine. Tokens-per-second throughput scales more linearly with memory bandwidth with this model, and so the Mac Studio delivers a 1.8X increase in throughput over GB10 and a 2.26X increase over Strix Halo.

But prompt processing speeds are only in line with (or even slightly worse than) Strix Halo here, which isn’t a fantastic result for latency if that’s important to your application.

Finally, we’ll look at gpt-oss-120b. Prompt processing again lands between Strix Halo and GB10, and tokens-per-second throughput is far higher than with either of those systems. But the Mac Studio’s tokens-per-second throughput is, on average, 1.5x higher than GB10 and 1.6x higher than Strix Halo, despite its much higher memory bandwidth. Those are still impressive leads, but they show that memory bandwidth is just one component in judging the performance of an inference pipeline.

On the whole, our LLM inference testing shows that the M4 Max GPU can generate many more tokens per second than competing systems with large unified memory pools, but a memory bandwidth specification alone isn’t a good proxy for delivered performance. The actual boost you get from the M4 Max in LLM inference is highly dependent on the architecture of the model you want to run, so real-world benchmarking is still important.

Image generation performance

LLMs aren’t the only AI workload that one might want to run locally. Image and video generation also matter, and the performance of these diffusion models is generally bound by the compute throughput of the underlying GPU.

Running our Flux.2 Klein image generation test on the Mac Studio initially failed due to an apparent lack of GPU support for the FP8 data type that the ComfyUI workflow uses by default.

We were able to modify the ComfyUI workflow to use the base FP16 version of Flux.2 Klein on the Mac Studio, but that initial compatibility issue shows that you might not enjoy as smooth an image or video generation experience on Apple Silicon as you might on GB10 or even Strix Halo.

Compatibility issues aside, the M4 Max GPU requires even more time than the already pokey Radeon 8060S in this image generation test. Nvidia has repeatedly demonstrated the DGX Spark running as an image and video generation companion for Macs in its demos this year, and this result demonstrates why that pairing might make sense.

Image generation performance with the latest M5 GPU and its dedicated Neural Accelerators for matrix math might be better here, but we don’t have an M5-powered system at hand to compare against the M4 Max. Apple does claim that the M5 Max is “up to 3.8x faster” at image generation than the M4 Max, so take heed if this workload is important to you.

General CPU performance

We already gauged the CPU performance of the M4 Max in our initial review, but we wanted to stack it up against Strix Halo and GB10 directly to see how it performs against these popular platforms in the AI space. To establish a CPU performance ballpark for the M4 Max, we fired up the just-released Geekbench 7 and got to it.

In this benchmark, the M4 Max turns in a whopping 27% advantage in single-core performance and a 21% boost in multi-threaded performance versus the Ryzen Max+ 395’s 16-core, 32-thread Zen 5 x86 CPU.

And comparing Arm to Arm, the M4 Max delivers a 39% single-threaded advantage over the Cortex-X925 “big” core on the Nvidia GB10, along with a 22% advantage in multi-threaded work.

Those results are seriously impressive on their own, but Apple has even more advanced CPU cores in its arsenal now on board the M5 series. We’d expect even larger advantages in general use from those chips.

Geekbench is fine for a quick synthetic test, but I wanted to make sure that its results held in a real-world task, so I compiled the CPU-only version of llama.cpp from source across all three of our test systems using the Clang 21 compiler running across all available physical cores.

With all 16 of its CPU cores in the game, the M4 Max finishes this compilation job in about 37 seconds, anywhere from 33% to 37% faster than the Nvidia and AMD competition here. That’s even faster than the M4 Max’s Geekbench multi-threaded performance delta would suggest.

At least going by these high-level tests, Apple’s CPU performance among SoCs with large unified memory pools is best-in-class, and by a wide margin. And again, that’s with the last-gen M4, not the latest M5.

When judging the efficiency of Apple Silicon or any other platform, one might be tempted to just glance at a power meter like a Kill-a-Watt with the system under load and leave it at that, but instantaneous power draw isn’t the right metric when judging the efficiency of a system for a given task.

Task energy (the amount of power consumed over time) is much more informative. A system that needs a lot of power but finishes a task quickly may consume less task energy and is therefore more efficient than one that consumes a moderate amount of power but takes much longer to finish a job.

To judge inference efficiency, we set up a fixed-length workload with gpt-oss-120b using llama.cpp and llama-benchy: first processing a 2048-token prompt against a 32768-token context depth, then generating 1024 output tokens. We chose that context depth to ensure that our test ran long enough to collect useful data. We recorded this data in watt-hours so that it reflects both energy consumed and time to completion.

At least in this synthetic workload, the M4 Max is more efficient than Strix Halo, but the sheer prompt-processing speed of GB10 lets it finish much faster and use less total power than the Mac Studio does, even though it’s outputting fewer tokens per second in the decode stage.

This test once again suggests that fixating on tokens-per-second throughput alone is perhaps not ideal for judging a system’s inference performance. An ideal LLM inference system would have both high prompt processing speeds and high tokens-per-second throughput for the shortest possible inference turn.

We’re curious how Apple’s latest M5 Max GPU would perform on this workload, as the company claims big gains on prefill compared to the M4 Max and has given that chip even higher memory bandwidth than the M4 Max. If that’s true, the combination of fast prefill and high tokens-per-second on decode would likely make the M5 Max competitive with GB10, if not an outright leader here.

We also gauged idle and load noise levels of these systems to see what it’s like to share space with them. For reference, the noise floor in our testing environment is about 32.6 dBA. We measure sound pressure levels at a distance of 18” from these systems to approximate what it’s like to work with them on top of a desk.

The Mac Studio is impressively quiet both at idle and under load. It barely raises the noise level in our testing environment at idle, and under load, its fans are barely perceptible. Practically any other noise you make or hear in your space will be louder. This is superlative performance.

Our time with a 128GB M4 Max Mac Studio demonstrates why these systems have been so popular among local AI trailblazers. For starters, the spec we tested has an incredibly potent 16-core CPU that’s leagues ahead of both Nvidia’s GB10 and AMD’s Ryzen AI Max+ 395 in both single-threaded and multi-threaded performance.

The massive memory bandwidth advantage of the 40-core GPU configuration of the M4 Max versus Strix Halo and GB10—546 GB/s vs 273 GB/s—also translates directly to higher tokens-per-second throughput in local LLM workloads. The M4 Max’s LLM inference performance is by far the highest of any unified memory platform we’ve tested.

But our tests demonstrate that there’s much more to delivered performance than that raw bandwidth figure. Depending on the model you choose, you might see the M4 Max generate as few as 27% more tokens per second or as many as 82% more TPS than an Nvidia GB10 box, at least with the llama.cpp inference engine.

This result emphasizes that you can’t just eyeball memory bandwidth and conclude that twice the bandwidth means twice as many tokens per second for LLM inference. Testing actual workloads remains essential.

And the Mac Studio doesn’t totally wipe the floor with the competition in our tests. For work like image generation, the Nvidia GB10 GPU still holds a large advantage over the M4 Max. Apple claims this workload will perform significantly better on the most recent M5 Max with its latest GPU architecture.

Beyond its impressive raw performance with LLMs and in general processing tasks, the Mac Studio is a master class in industrial design. It looks distinctly classy on a desk, and its cooling design keeps it incredibly quiet. Under load, it’s barely louder than the Nvidia GB10-powered Dell Pro Max and Ryzen AI Halo are at idle. And when the Studio itself is idling, it’s practically silent. This level of performance in such a compact, quiet, and refined chassis is incredibly impressive.

There’s one major asterisk, though: the ongoing RAMpocalypse means that even if you do want to use a Mac Studio as a foundation for your local LLMs, these systems are no longer available in the configuration that we tested.

A Studio with the high-end M4 Max as seen in this review currently tops out at 64GB of RAM for $5299, and the M3 Ultra version tops out at 96GB (with an even higher 819GB/s of memory bandwidth) for the same price. But neither of these configurations are available from stock, and lead times for both are about two months out, meaning that you practically can’t buy a Mac Studio new right now.

If you really want to try out local AI on Apple Silicon with a large memory pool to match, the best current option may be a MacBook Pro with an M5 Max inside, which is still readily available from stock, still allows you to get up to 128GB of RAM, and benefits from Apple’s latest GPU architecture with dedicated Neural Accelerators for matrix math operations.

But depending on your SSD choice, a 128GB M5 Max MacBook Pro with the highest-end 16-core CPU and 40-core GPU is still a $7K to $10K system today. Even in today’s world of rising prices for access to frontier models, that’s a lot of tokens from the cloud – and, almost assuredly, more intelligent ones – than anything you can get from a single 128GB system from any vendor.

But if you’re dead-set on a single Apple Silicon system with the most memory possible, and you want it now in order to ride the local AI wave, a maxed-out M5 Max MacBook Pro is your only ticket at the moment. Given the refinement and flexibility of the Mac Studio as a desktop or server, we can only hope Apple brings that chip to this system, too.

Following price hikes on the Mac and iPad due to the ongoing memory and storage crisis, Apple has announced a new leasing program, Apple Upgrade. The new offering will be "provided" by Klarna and available in Apple Stores in the U.S., as well as the Apple Store online and the Apple Store app.

There will be 24- and 36-month leasing options for the Mac and iPad, while the iPhone and Watch will have 12- and 24-month options.

Mac leases will start at $24.99 per month, which includes the MacBook Air, MacBook Pro, iMac, and Mac Studio. Apple's announcement leaves out its cheapest systems, the MacBook Neo and Mac Mini.

iPad leases start at $11.99 per month and include the iPad Mini, iPad Air, and iPad Pro, but not the base-level vanilla iPad.

It's a similar pattern for the Apple Watch, with leases starting at $11.99 but leaving out the Apple Watch SE 3. But the iPhone lineup (leases start at $17.99) doesn't have any exceptions, including the iPhone 17, 17e, and 17 Pro and Pro Max, and iPhone Air.

These are not purchases or loans. Customers can sign up for Klarna leases in-store or online (including a soft credit inquiry that doesn't affect your credit score). At the end of a lease, you can return your device and start a new lease to upgrade, purchase the device with a "one-time payment," or return the device and exit the lease entirely.

Unlike the iPhone Upgrade Program, Apple Upgrade leases don't include AppleCare, but Apple will let you buy a subscription. AppleCare also. saw some price hikes, likely because the components that are serviced under the extended service program are more expensive.

In fact, Apple Upgrade will replace the iPhone Upgrade Program in the United States, likely upsetting some diehard fans. Apple will allow those customers to lease a device through Klarna, buy the phone outright, or finance with monthly installments on their Apple Card. As always, there's carrier financing for phones.

Following unprecedented price hikes on the Mac and iPad, Apple is now working on a leasing program to alleviate customers during the AI rush. According to a new Bloomberg report, the program is called "Apple Upgrade" and will allow you to lease a midrange or high-end device for up to 36 months, at the end of which you can choose to keep it. Apple is purportedly partnering with Klarna on the service.

Apple Upgrade will reportedly be a subscription service with monthly payments spanning across 24 months for iPhone and Apple Watch, or 36 months for Mac and iPad, including the Pro models. According to the report, Entry-level hardware like the MacBook Neo is not included in this program. Education or business discounts are also excluded; Apple is explicitly targeting regular consumers who've been affected by the ongoing component crisis.

According to Mark Gurman, customers will be subject to a soft credit check and only be approved for the program after passing it. The program will be available both in-person at physical Apple Store locations and online on Apple's website. At first, it will only be available in the United States, but we don't have any info on whether Apple Upgrade will expand beyond the region later.

As mentioned, Klarna is reported to be the official broker between Apple and the customer. The leasing structure would closely follow standard practice, where you can choose to keep the asset at the end after paying a residual fee, or return it, in which case you were just renting the hardware. Customers will also reportedly have the option to upgrade to a newer model if they finish their payments early.

For instance, if you leased an iPhone 17 Pro Max but Apple announced the 18 Pro Max a few months later, you could simply get the new device early if the old one was already paid off. Some of you will know that Apple already has an upgrade program for iPhone, but this one is different since it expands to Mac, iPad, and other accessories, and, more importantly, it doesn't come with AppleCare.

In fact, Bloomberg says Apple will eventually phase out the iPhone Upgrade program in favor of Apple Upgrade as a whole. New enrollments for standard financing through the Apple Card will also reportedly be discontinued. The company will advertise lower monthly payments to attract customers, which will ultimately allow it to keep a stable income during one of the most turbulent consumer tech eras the world has ever seen.

TSMC intends to raise base quotes on advanced chip production services by up to 10%, according to Nikkei, which cites people with knowledge of the matter. The price hike reflects increased demand for sophisticated processors by the AI sector, raising costs of tools and materials, as well as amplified investments in new production capacities.

For advanced process technologies — which TSMC considers 7nm-class and below — TSMC plans to raise baseline prices by 5% to 10%, depending on the particular production node and customer, the report claims. Furthermore, customers that need additional HPC chip capacity beyond their original volume requirements will reportedly have to pay another 10% to 15% premium on top of the standard increase, which means that some services will get a price hike of around 25%, if the report is accurate.

TSMC also intends to increase prices for mature manufacturing technologies, including its 12nm, 16nm, and 28nm-class nodes as well as other legacy fabrication technologies, the report claims. Increases could reach 10%, although certain nodes will reportedly see smaller adjustments, according to Nikkei.

Advanced technologies generated around 77% of the foundry's revenue in Q2 2026, whereas mature nodes accounted for 23%, which essentially means that TSMC is hiking prices on all of its services.

The company reportedly began discussing the new pricing with customers around June and completed negotiations in July. Rather than introducing higher rates immediately, TSMC opted to implement them from the beginning of 2027 to give clients like Apple, AMD, Nvidia, and MediaTek additional time to accommodate the changes and adjust their prices accordingly.

Since TSMC produces the lion's share of advanced processors for AI, HPC, networking, and smartphone applications, its price hikes will inevitably create a ripple effect in the industry and will make almost all electronics more expensive.

TSMC is not alone in raising prices these days. Vanguard International Semiconductor has also raised prices, while UMC began implementing increases in July. Also, memory makers have increased prices significantly, making TSMC management jealous. Intel also recently increased prices of its client and data center CPUs, citing market demand.

"I am really jealous about memory companies' 86% gross margin," said C.C. Wei, chief executive of TSMC, during the company's earnings call with financial analysts and investors. "86% [margin at memory makers] – 68% [margin at TSMC], I will be happy about that."

TSMC rarely comments on its prices to a large degree because they vary based on volumes and relationship with a particular client. Nonetheless, the head of the company stressed that the company has no intentions to increase prices suddenly or dramatically.

"So we do not suddenly increase our price by which I like to have 4x or 5x," Wei said. "You cannot survive for that kind of... for your customer to survive for that kind of price increase. So we earn our value, and we make sure that our profit, our gross margin, is enough for our long-term sustaining expansion, that is to the benefit of my customers and TSMC also, that is our philosophy."

There's no denying how good the build quality and construction of Apple's MacBook Pro lineup are. They are what every other high-end laptop manufacturer tries to emulate, which lets Apple charge a very high price for the privilege of owning one. If you're looking for a high-performance laptop for coding, video production, or even some light LLM/AI tasks, then this deal on the 14-inch Apple MacBook Pro with M5 chip for $2,699 is worth a look. B&H Photo has slashed $400 off the $3,099 list price to bring it down to this price, and give you the chance to pick up a powerful laptop with 48GB of memory.

● Check out this deal at B&H Photo

This 14-inch MacBook Pro uses the latest M5 Pro chip, with a 15-core CPU and 16-core GPU with a Neural Accelerator in each core, delivering amazing processing power in a small package. This model sports a hefty 48GB of unified RAM, a 1TB SSD for storage, and a crisp 3024 x 1964-pixel Liquid Retina XDR display (1600 nits peak HDR brightness), with superb battery life (up to 24 hours).

Other features include a 12MP Center Stage camera with Desk View support, an in-house N1 networking chip that combines Wi-Fi 7 and Bluetooth 6, an SDXC Slot, Thunderbolt 5, and a MagSafe 3 connector.

We had the chance to review the MacBook Pro 14-Inch and found that the M5 chip delivers solid performance gains over its M4 predecessor. The SSD performance has greatly improved, and the rocking speaker system is still a cut above similar laptops from competitors. In our benchmark testing, the MacBook Pro's M5 CPU delivered between a 10 to 20% improvement over the prior generation's M4 chip in CPU performance.

If the price tag of $2,699 and 48GB of RAM are a little too high for your tastes, you can also pick up the same 14-inch MacBook Pro with M5 Pro chip, but with 24GB of RAM for $2,299, with the same hardware and features as our main deal.

If you're looking for more savings, check out our Best PC Hardware deals for a range of products, or dive deeper into our specialized SSD and Storage Deals, Hard Drive Deals, Gaming Monitor Deals, Graphics Card Deals, Gaming Chair, Best Wi-Fi Routers, Best Motherboard, or CPU Deals pages.

You can also join the Tom's Hardware deals Discord for up-to-the-minute hardware deals.

Apple's planned M7 Ultra chip is being designed to support up to 1.5 TB of unified memory and to push AI performance toward the class of Nvidia's Blackwell accelerators, according to a new Bloomberg report published by Mark Gurman. But whether the lofty top memory config can ship at all will depend on the state of the memory market, and the part isn't expected until 2028. The same report says Apple has compressed its Mac silicon timeline, taping out the M7 roughly six months after the M6.

We've already heard that Apple plans to release a base M6 chip this fall for entry-level Macs, then skip the Pro, Max, and Ultra versions of that generation and move straight to the M7 line. However, Gurman now reckons that we'll see a base M7 in the first half of 2027, M7 Pro and M7 Max at the end of 2027, and the M7 Ultra in 2028. Apple reportedly began taping out the M7 about six months after it started the same process for the M6, which is what has enabled the company to pull the schedule forward.

The 1.5 TB target for the M7 Ultra is roughly twice the capacity Apple has planned for the M5 Ultra, per Gurman, who tied the configuration directly to memory availability. Apple already pulled the 128GB Mac Studio this year over supply constraints as DRAM prices climbed, and a 1.5 TB part would call for far more of the same scarce, high-cost memory.

Apple's current M3 Ultra already reaches 819 GB/s of memory bandwidth by fusing two Max dies, and it's the Ultra tier, not the base chips, that carries the heaviest local-AI workloads. Gurman describes the M7 Ultra as a large step up in AI performance rather than stated parity with Nvidia's data-center silicon. "I'm told the processor dramatically upgrades AI performance, bringing it closer to the class of dedicated AI accelerators such as Nvidia Corp.'s Blackwell," Gurman wrote in his report.

Apple is also preparing an AI server built on the M5 Ultra under the code name J246 for deployment soon, with a second server chip based on the M7 Ultra planned for 2029, according to the report. The 2028 generation, which includes an M8 chip code-named Soko and a high-end Mac part called Cardinal, moves to a 1.4nm process. That aligns with TSMC's A14 node, which the foundry has scheduled for mass production in the second half of 2028.

None of the dates or specifications have been confirmed by Apple.

Sotheby’s is preparing a blockbuster History of Science & Technology sale packed with amazing artifacts and collectors’ items. While there are many items that could justify being singled out, the famous auction house has chosen to spotlight a working sample of the Apple-1 computer serial number 01-0033 (1976) on its YouTube channel. Remember this first 50 hand-built-by-Wozniak batch has a place in history as the first commercial Apple computers ever sold.

In the video above you can see Harry Sadler, the original and current owner of the Apple-1 computer up for auction, preparing, booting and demonstrating this ancient ‘pre-built.’ Sadler underlines that his rare example was bought from the Byte Shop in 1976. While he ended up working in IT, Sadler said that it was mainly a hobbyist machine, for experimenting, tinkering with BASIC, and so on.

Interesting, Sadler took his Apple-1 back to the Byte Shop after a couple of weeks to get a minor repair done. There he met Steve Jobs and inquired about vacancies at Apple Computer. Nothing came of that because of admitted inexperience, but in 2010 Steve Jobs visited and bought the Siri Inc. startup, where Sadler was the 10^th employee. So, Sadler did actually end up working for Apple.

The estimate for this machine is $300,000 - $500,000. It comes with original accessories like a cassette deck and keyboard, vintage monitor. There’s also an original Apple-1 Operation Manual, In-house copy of Apple Basic Manual, Apple-1 Cassette Interface Manual, Microchess manual, and various software on cassettes

I think this is rather high, as one of the rare wooden case models known as the ‘Moore Apple-1’ only achieved $475,000 (w/BP) for its client last year.

Other notable computers in the sale

This Sotheby’s sale has quite a few other notable computers among the tech artifacts listed. For example, there’s the Steve Jobs stage-used NeXT Computer from the ‘NeXT Introduction’ in 1988, and a similar model for European demos. Then there’s a Mark-8 minicomputer from 1974, a MITS Altair 8800 computer from 1974, as well as some complete sealed vintage Mac systems from early on in Apple’s history.

On the more scientific side of things, there’s lots of interesting artifacts associated with, owned, or used by iconic individuals like Thomas Alva Edison, Alexander Graham Bell, Richard Feynman, and even Albert Einstein.

The History of Science & Technology takes place on July 15, starting at 14:00 EDT live in Sotheby’s saleroom at 945 Madison Avenue in New York.

Apple filed a federal lawsuit against OpenAI on Friday, accusing the AI company and its chief hardware officer of stealing its trade secrets.

"OpenAI and its cohorts, led at least in part by former Apple employees, have recruited candidates from Apple, extracted their knowledge of Apple’s sensitive and confidential information, and then continued to exploit that knowledge once they arrived," the complaint reads. "As a result, OpenAI has misappropriated Apple’s trade secrets and confidential information in a variety of ways."

The suit, filed in the Northern District of California, names OpenAI technical staff member Chang Liu, chief hardware officer Tang Tan, OpenAI, and io Products as defendants. The last of that group is notable because it was founded by Tan in collaboration with former Apple design head Jony Ive, Evans Hankey (Ive's successor at Apple), and former Apple designer Scott Cannon. Notably, the complaint seems to attempt to avoid naming the founders, though Ive's name is cited in a URL.

Tan previously served as a vice president of product design at Apple, working on the iPhone, AirPods, and Apple Watch. Liu served at Apple as a senior electrical engineer.

In the complaint, Apple alleges that it reached out to OpenAI in February with concerns, but that OpenAI did not respond. Apple claims that Tan attempted to gain secrets from Apple employees, including asking prospective job candidates to bring components for "show and tell" sessions and used his knowledge of the company to squeeze more information out of candidates. The suit claims that Liu never returned a company laptop, and used an authentication bug to access Apple files.

Apple also claims that OpenAI told incoming employees how to leave their former job, suggesting they stay as long as possible and not disclose their former employer in order to continue to access confidential information.

"At every level, from members of its Technical Staff to its Chief Hardware Officer, and in coordination with business partners, OpenAI has been stealing Apple’s trade secrets and confidential information," the suit reads. "As a natural result, OpenAI’s nascent hardware business now rests on the shakiest of foundations, rotten to its core by its illegal reliance on misappropriated trade secrets."

OpenAI did not immediately respond to a request for comment from Tom's Hardware. Apple's lawsuit claims that over 400 former Apple employees currently work at OpenAI.

Apple is rumored to be working on a number of AI-powered hardware projects, including AirPods with cameras, a pendant, and home robots. It's less clear what hardware OpenAI may be working on, though The Information suggested the company has a HomePod-style smart speaker in the works.

Apple is requesting a jury trial, damages, attorney fees, and orders that the OpenAI may not use Apple's trade secrets, among other injunctions.

In May, Bloomberg reported that OpenAI was considering legal action against Apple because it expected deeper integration and more users from ChatGPT features built into iOS.

If the trial does go to court, it's sure to be a dramatic one, potentially dragging several former high-level Apple employees into testimony through discovery and testimony.The trial, Apple Inc. v. Liu et al, is case 5:26-cv-07078 in the United States District Court in Southern California.

As the ongoing component crisis worsens, even the biggest companies are no longer shielded. Apple, in a historic first, raised prices on a fleet of its products yesterday, saying it could no longer protect its customers from the soaring memory and storage costs. Now, the Financial Times is reporting that Apple is trying to lobby the U.S. government to secure official clearance to buy (cheaper) memory from China's CXMT.

CXMT is not outright banned by the White House, as it's not on the Defense Department's "Entity List," which prohibits Chinese companies entirely on national security grounds. Instead, it's listed in the 1260H list, which still designates it as a Chinese military company, and doing business with it would cause major reputational damage to an American company.

Apple has apparently been reaching out to its allies in Washington to persuade the government to allow it to purchase RAM from CXMT. Unlike Micron, Samsung, and SK Hynix — the coalition controlling 90% of the world's DRAM — CXMT has no real incentive to chase AI buildouts. It can therefore provide RAM at much more reasonable prices without compromising performance.

Recently, Corsair Vengeance DDR5 sticks from China were found to use CXMT modules underneath, signaling the company's entry into the mainstream. That was a 6,000 MT/s CL30 kit, so it seems CXMT certainly has the ability to produce modern silicon at a scale sufficient for a major PC vendor like Corsair. OEMs such as Dell and HP are also adopting Chinese-made RAM for those region-bound systems.

Before that, CXMT has already demonstrated its capabilities for cutting-edge DRAM manufacturing; late last year, it showed off production-ready DDR5-8000 and LPDDR5X-10667 modules. Volume is another issue entirely, but Beijing has certainly figured out how to make performant RAM. If a buyer like Apple were to enter the picture now, even just for the Chinese market, it would completely shift the paradigm.

Of course, such a precedent is subject to approval, and high-ranking officials seem adamant against it. John Moolenaar, the chairman of the House China committee, has warned that Apple partnering "with a Chinese military company would be a grave mistake." He told Financial Times that it's imperative for America not to rely on foreign supply chains for something as critical as DRAM.

Apple is in a dire situation itself, having lost $265 billion off its market cap the day it announced price hikes on MacBooks and iPads. In an uncharacteristic move, Tim Cook even admitted he'd never seen anything like this in his 40+ year stint. The Cupertino giant has always enjoyed comfortable margins on memory and storage upgrades, but it can no longer dam the "hundred-year flood."

"Trump can show the courage to keep American memory alive for our security and our competitiveness or pour it down the drain so Tim Cook can squeeze out a few more points of margin," remarked an ex-official. Another security expert said it wouldn't make sense for the administration to fiercely protect critical earths and rare minerals, while suddenly conceding in the AI race (which is the reason behind the RAMpocalypse).

Whether Apple's efforts in Washington succeed or not is also predicated on the changing of the guard, since Tim Cook will hand over the baton to John Ternus come September — a logistics wizard leaving at perhaps the most decisive moment he was needed. The AI boom is here to stay, and so is the component crisis for now, so it remains to be seen how the company navigates the oncoming challenges.

Apple will release a base M6 chip for entry-level Macs this year but skip the Pro and Max versions of that generation, jumping instead to an accelerated M7 family built around on-device AI, according to a report from Bloomberg. It’s the first time since the M1 that a generation would arrive without higher-end variants, and it pulls the M7 forward by as much as six months, with Pro and Max parts now due in late 2027 and an Ultra in 2028.

The M6, codenamed Komodo, is set for entry-level machines, including a refreshed 14-inch MacBook Pro, which has seen unprecedented price rises to $1,999. According to the report, which quotes individuals who asked not to be named, it reaches around 200 GB/s of memory bandwidth against 153 GB/s on the M5, with a redesigned GPU carrying up to 12 cores, up from 10. Every family from the M1 through the M5 paired its base silicon with Pro and Max derivatives, and three of them also gained an Ultra.

The base M7, codenamed Delos, targets roughly 240 GB/s and could come in the first half of 2027, with the Pro, Max, and Ultra parts grouped internally under the Andros codename. Apple is fast-tracking the line to meet demand for on-device AI and heavier graphics work. Apple has declined to comment, and none of the specifications or dates have been confirmed by the company.

Memory bandwidth dictates how fast a chip can move the large data blocks, the main bottleneck for AI inference, and the climb from 153 GB/s to 200 GB/s to a planned 240 GB/s is a roughly 56% bandwidth increase from M5 to base M7. Those are figures for the base parts. The current M3 Ultra already delivers up to 819 GB/s by fusing two Max dies, which is why the high-end tiers, not the base chips, carry the heaviest local-AI workloads. The M6 also reportedly pairs an upgraded Neural Engine with faster GPU and CPU cores.

Apple is still planning to release an M5 Ultra this year, codenamed Sotra, with around 36 CPU cores, 80 GPU cores, and tested support for up to 768GB of unified memory. It would refresh a Mac Studio that currently runs the M3 Ultra, after Apple skipped an M4 Ultra altogether.

That 768 GB ceiling sits against a current M3 Ultra Mac Studio capped at 96 GB, after Apple pulled its 512 GB tier and raised the 96 GB-to-256 GB upgrade to $2,000, then dropped the 128GB option as well. Whether the highest M5 Ultra memory tiers reach buyers depends on supply constraints; those haven’t eased and are unlikely to in the near future.

Apple raised prices across its Mac and iPad lines on Thursday, part of a crunch that outgoing CEO Tim Cook called a “hundred-year flood” in an interview with the Wall Street Journal, blaming AI server demand for pulling high-bandwidth memory away from consumer hardware.

Apple has just announced that it has hiked prices for its line-up of MacBooks and iPads. MacBook Air and MacBook Pro models are expected to see prices rise by up to $400 in some cases, while the budget-friendly MacBook Neo is expected to see its MSRP rise by $100, from $599 to $699.

Luckily, we're in Amazon Prime Day sales week, and discounts are live on Amazon for several Apple MacBooks from 2026 and 2025. Now really might be your last chance to secure a decent deal on a new MacBook before Apple's new pricing structure filters down, as we'll be expecting tech retailers (including Amazon) to raise prices to match Apple's own soon enough.

● Check out these Apple deals at Amazon

These deals won't be around for long, either. With price rises now confirmed, those thinking about purchasing one of these Apple laptops will be rushing to hit the purchase button before paying extra. Stock shortages, especially on the cheapest models, could become a real possibility by the end of this sales week.

If you're interested, here are some of the best MacBook deals on sale right now, because you'll want to jump on them while you can.

Prime Day Deals on Apple MacBooks

MacBook Pro, Air, and Neo models from 2025 and 2026 are all on sale right now at Amazon for Prime Day. Amazon hasn't put all models on sale, however, but here are some of the deals we've spotted so far.

One thing to point out, however, is the MacBook Neo. Apple has just raised its MSRP to $699, meaning that Amazon's $589.99 price is $110 cheaper. This will not last long, so make sure to grab it at this price.

Other MacBook models, including the MacBook Air 2026 with a 512GB SSD, are seeing $200 price rises. The MacBook Pro, meanwhile, is getting a price rise of up to $400. This isn't reflected in these Amazon deal prices yet, which have knocked up to $150 off Apple's original prices, so you're guaranteed an even bigger bargain today if you can get your order in quickly enough.

Apple MacBook Prime Day Deals

More Prime Day Tech Deals

Best Tech and PC deals | Best gaming PC deals | Best RAM combo deals | Best 3D printer deals | Best RAM deals | Best gaming laptop deals | Best monitor deals | Best Wi-Fi Router deals | Best GPU deals | Best SSD deals | Best hard drive HDD deals | Best CPU deals | Best gaming chair deals | Best PC building tool deals | Best PC peripherals deals | Best filament and resin deals | Best motherboard deals | Best CPU cooler deals | Best PC case deals | Best Dell and Alienware deals | Best USB charger deals | Best gaming and productivity laptop deals under $1,000 | Best laptop PC deals

Also, you can join the Tom's Hardware deals Discord for up-to-the-minute hardware deals.

Apple has made the unprecedented decision to hike the prices of all its current computers and tablets. iPhone prices remain as they were, for now. Bloomberg indicates that the rather steep price rises have been precipitated by increased costs of memory and storage. One of the most impactful hikes we see is the entry-level MacBook Pro going from $1,699 to $1,999. However, even the recently released but resource-starved MacBook Neo has had another $100 piled onto its price, so now it starts at $699 before tax.

In addition to Gurman’s highlighted price increases, we spotted that the cheapest Mac Studio M4 Max (mini PC desktop) has had its price increased to $2,499 from $1,999. There’s a lot to take in, so check out our at-a-glance before and after table for quick reference.

Bloomberg’s resident Apple expert, Mark Gurman, quoted an apologetic company rep. “We know this is not welcome news, and we are working tirelessly to find solutions,” a spokesperson from Apple told the reporter. Other interesting assertions from that conversation were that Apple had never seen component prices rise so high, so quickly. Moreover, its representative claimed that the firm did all it could to shield customers from any knock-on device price rises, stating the company had "shielded our customers from these increases so far, but we have now reached a point where we need to begin raising prices on a number of products including today’s increases for iPad and Mac."

While newsrooms might be agog at these price rises being implemented today, the writing has been on the wall of the RAMpocalypse for several months. Back in April, Tim Cook publicly forecast that the component shortages and price rises didn’t appear to be going anywhere soon. That statement came in the wake of the withdrawal of the entry-level $599 configuration of the Mac Mini, and Apple softening the blow of higher prices of its latest MacBook Air and MacBook Pro models by upgrading memory and other tech specs.

Until now, it seemed like Apple’s massive buying power and vertical integration had managed to help it hold off computer and tablet price increases. Today, the brakes came off, pretty dramatically.

For years, Apple kept its M1 MacBook Air around as the ultimate budget pick in its laptop lineup. Now, the M3 MacBook Air from 2024 has just hit the same heady discount heights, plunging to $799 at Amazon in the early Prime Day sale in a deal that makes the MacBook Neo look pretty tame by comparison.

• Check out this deal on Amazon

The M3 MacBook Air came out in 2024, so it's a couple of years old now. However, if you want a fantastic portable machine that will give you great battery life, this is not to be sniffed at. For just $200 more, this blows the MacBook Neo out of the water. Its M3 chip has an 8-core GPU and 10-core CPU, and you also get 16GB of RAM and 512GB SSD. That means you get a better CPU and GPU core count, twice the RAM, twice the storage, and higher memory bandwidth versus the Neo. The M3 also offers drastically better battery life and a host of other extra features.

Both the M4 and M5 MacBook Air are on sale for around $950 at Amazon right now. That makes this deal $150 less than those more expensive options. While the M4 and M5 chips are faster than this M3 model, this older version is no slouch.

Our Geekbench testing does reveal a fairly substantial uplift on the M4 model, which scored around 3,800 in similar Geekbench 6 single-core tests and nearly 15,000 in multi-core.

What you're getting here is really the ultimate budget MacBook right now. Of course, the MacBook Neo is cheaper at $589, but for that price, you get a meager 8GB of RAM and just 256GB of storage, half what you get with the M3. The M3 has more CPU cores (10 vs 6) than the MacBook Neo, and more GPU cores (8 vs 5). Battery life is also rated for longer (18 hours versus the Neo's 16 hours), and the M3 comes with significantly better 102 GB/s maximum memory bandwidth compared to the Neo's 60 GB/s.

Versus the Neo, the M3 MacBook Air also has a higher resolution display that is slightly larger, and has better color with Wide color (P3) and True Tone Technology. The M3 also has more speakers, support for spatial audio, better microphones for calls, and support for high-impedance headphones, which you won't find on the MacBook Neo.

If you can stretch from the Neo to this $799 Macbook Air, this one is a no-brainer as a great laptop for a college student, a gift for a relative, or a high school student.

Join the Tom's Hardware deals Discord for up-to-the-minute hardware deals.

Best Tech and PC deals | Best gaming PC deals | Best RAM combo deals | Best 3D printer deals | Best RAM deals | Best gaming laptop deals | Best monitor deals | Best Wi-Fi Router deals | Best GPU deals | Best SSD deals | Best hard drive HDD deals | Best CPU deals | Best gaming chair deals | Best PC building tool deals | Best PC peripherals deals | Best filament and resin deals | Best motherboard deals | Best CPU cooler deals | Best PC case deals | Best Dell and Alienware deals | Best USB charger deals | Best gaming and productivity laptop deals under $1,000 | Best laptop PC deals

Apple CEO Tim Cook has warned in an interview with the Wall Street Journal that price hikes on Apple products are unavoidable because the price of memory has increased to a degree that the company must now pass on increases to customers. The outgoing chief executive of Apple did not disclose the scope of the increases or when to expect them to happen, though the warning itself is noteworthy.

Cook said in the interview that price increases had become necessary due to skyrocketing prices of LPDDR and 3D NAND memory, which the company uses in its PCs, smartphones, tablets, and other products. He noted that Apple had attempted to offset rising component costs and protect customers from higher prices, but indicated in the interview that the company could no longer absorb the increases indefinitely. While Cook declined to discuss timing or the magnitude of the planned price hikes, some Apple products may see higher prices sooner rather than later. The company already raised the base price of its Mac Mini last month and eliminated its highest-end model.

TechInsights estimates that Apple will need to hike the price of a flagship iPhone Pro model by about $270 to maintain its current gross margins, which could push flagship iPhone pricing into a substantially higher price band. To make matters worse, Apple faces rising DRAM requirements as it expands memory-hungry on-device AI capabilities.

Apple is a unique company in the sense that it is both one of the world's largest suppliers of smartphones and one of the industry's biggest PC OEMs. Still, its Mac business is negligible compared to the iPhone business.

Memory makers are more inclined to ship non-volatile NAND memory to smartphone makers as they can ship LPDDR with it, according to Nelson Duann, a senior vice president of Silicon Motion, speaking in an interview with Tom's Hardware. If Duann is correct that NAND makers prefer smartphones because they can bundle NAND and LPDDR sales, then Apple should be among the most favored customers in the industry, not among the disadvantaged ones.

Also, keep in mind that historically Apple has used long-term supply agreements, prepayments, equipment financing, and advance capacity reservations to secure key components, including DRAM, NAND flash, displays, advanced packaging, and even semiconductor foundry capacity. As one of the world’s largest electronics manufacturers and semiconductor buyers, Apple is among the few companies capable of negotiating with memory suppliers from a position of considerable strength.

As a result, Apple is not in a situation where it cannot get enough memory; it can probably get more than other suppliers of smartphones and PCs, but it surely has to buy both DRAM and NAND at a premium. That said, it is not surprising that Apple will have to increase prices; what remains to be seen is the magnitude of the increase.

President Donald Trump said on Thursday that Apple has agreed to work with Intel to “design and build” chips in the United States, posting the claim to Truth Social before either company confirmed any arrangement. Apple and Intel didn’t respond to initial requests for comment from Reuters, and neither has issued a statement acknowledging a finalized deal.

Trump says the deal is part of his administration’s push to reshore semiconductor manufacturing, writing that the U.S. needs to build its chips domestically and that he "decided to help Intel because we need to design and build our Chips right here in America.” He credited earlier government intervention for drawing Nvidia and Elon Musk's TeraFab project to Intel's foundry.

Apple already designs its own silicon and has done so since it dropped Intel processors from its products in 2020. Any arrangement here would purely be a foundry deal, with Intel acting as a contract manufacturer for chips Apple designs in-house. It wouldn’t return Apple to Intel-designed processors, and it certainly wouldn’t displace TSMC from Apple's flagship products.

The logic behind any such deal for Apple would be simple supply diversification; the company remains dependent on TSMC, whose leading-edge capacity is being absorbed by AI customers, including Nvidia and AMD. Placing a lower-volume part on a second source reduces single-foundry exposure without touching the iPhone and high-end M-series lines that rely on TSMC's most advanced nodes.

We’ve known about much of this for a little while now. Reporters caught wind of a preliminary deal last month, following discussions that ran for more than a year. Before that, analyst Ming-Chi Kuo and GF Securities outlined the likely scope: Apple's M7 SoC built on Intel's 18A-P process, powering the MacBook Air and entry-level iPad Pro, with mass production targeted for late 2027.

Those two product tiers accounted for roughly 20 million units in 2025, with annual shipments expected to settle between 15 million and 20 million — significant for a new foundry customer, but small enough to leave TSMC's revenue mix intact. Reporting has also suggested that Apple's A21 iPhone chips will move to Intel's 14A node around 2028, but none of this has been confirmed by Apple or Intel.

Trump's announcement comes just two days after Intel said its 18A-P process had entered risk production, announced at the VLSI Symposium in Honolulu. The node is the first performance-enhanced version of 18A, and Intel claims it delivers 9% higher performance at the same power, or 18% lower power at the same performance, while staying design-rule compatible with 18A so customers can reuse existing IP. CEO Lip-Bu Tan told investors last month that he expects multiple foundry commitments to close in the second half of 2026.

Intel has yet to confirm a single leading-edge customer for 18A, and any commitment from Apple would be a best-case scenario in terms of securing the validation the company both desperately wants and needs. The U.S. government holds a 10% stake in Intel, and its shares rose as much as 9% in premarket trading on the news, extending a run that has lifted the stock 464% over the past 12 months to a market cap of $608.7 billion.

In a turbulent conflict-shaken world rife with geopolitical rivalry, tech export bans seem increasingly prevalent. However, history provides some context, and some fun in this case, as industry watchers rightly highlight that computer technology export bans are nothing new. Here we look back at the Apple Power Macintosh G4 export ban from 1999, and Steve Jobs making marketing gold from the situation.

In the summer of 1999, the U.S. government put export restrictions on Apple’s newly launched Power Mac G4 desktop tower systems. Being capable of a claimed performance of “over 1 billion calculations per second” meant these stylish ‘graphite’ translucent designs were “classed as a weapon” and thus routinely banned from export to 50 nations worldwide.

Though Apple naturally sought to get the restrictions lifted, according to contemporary reports, its new interim chief executive officer, Steve Jobs, didn’t miss a great marketing trick. Check out the resulting 30-second nugget of marketing gold in the video we have embedded.

“For the first time in history a personal computer has been classified as a weapon by the US government,” narrates the voice over, backed by the theme tune from the classic war movie The Great Escape. “With the power of over 1 billion calculations per second the Pentagon wants to ensure that the new Power Macintosh G4 does not fall into the wrong hands.” Jobs didn’t miss a chance to belittle Intel with the closing line “As for Pentium PCs, well they're harmless.”

As a reminder, the first Power Mac G4 machines, codenamed Yikes!, did indeed wield envious compute power at the time. They were marketed as offering up to double the performance of their G3 predecessors and triple that of Pentium III PCs clock-for-clock. Low End Mac recalls “The 400 MHz Yikes! (a.k.a PCI Graphics) offers 0.8-3.2 gigaflops (billion floating point operations per second) performance; by government definition in 1999, it was a supercomputer.”

The 400 MHz model was the entry-level model at launch, so as well as it being banned from export the faster 450 MHz and 500 MHz were also off the menu if you lived in one of the countries on the export control list at the time. Bill Clinton’s administration raised the gigaflop threshold for U.S. export controls in January 2000. Then, with the threshold raised to 6.5 GFLOPS, Apple could again begin unrestricted G4 exports.

History repeating

The latest tech export controversy headlines revolve around the recently launched version of the Anthropic Claude AI model, dubbed Fable. We have also widely reported on hardware bans, with Nvidia GPUs probably the most notable example. However, western governments have also sought to restrict exports of semiconductor tools and software to places like Russia, China, Iran, North Korea etc.

Apple introduced its next major macOS release, macOS 27 Golden Gate, at its WWDC event. The new operating system update, coming this fall, includes a series of new artificial intelligence gestures, as well as iterations to the "liquid glass" design introduced in last year's release, macOS 26 Tahoe.

Craig Federighi, Apple's senior vice president of software engineering, said that the 27 releases were focused on more polished and intuitive operating systems, trust and safety (including for children), and updates to Apple Intelligence and Siri. Unlike previous years, Apple spent most of its time highlighting changes that affect the entirety of its platforms, meaning that many of the improvements on Mac are also available in some form on the iPhone or iPad, and vice versa, reflecting the continuing deep integration in Apple's ecosystem of products, including iPhone (iOS 27), iPad (iOS 27), and watchOS (watchOS 27).

macOS will include some specific upgrades, including ultrawide display support with higher resolutions, and an updated video podcast player, while many others, like updates to iCloud shared albums and changes to Maps, are across multiple platforms. In many ways, the focus on fixes is reminiscent of 2009's OS X Snow Leopard, which was famous for fixing problems rather than introducing tons of new features.

Liquid Glass

Apple is making changes to Liquid Glass across its platforms to make content more readable and decrease distractions. Glass will now better diffuse content behind it.

A new slider in settings will let you move from from ultra-clear to fully tinted, letting you customize how much it affects readability.

Specifically on the Mac, Apple is adding a uniform tool bar across the top of apps, which harkens back to more traditional Mac design. Expanded sidebars will move to the edge of windows, and sidebar icons will regain their colors. Additionally, every window will have tighter corner radii, even if they're not updated, for improved consistency across the OS. These were all complaints from Mac diehards over the last year.

In app icons across platforms, there will be additional layers of glass in icons to make them sharper and more defined.

Improvements

Apple said that it is bringing a massive amount of improvements under the hood of macOS and other platforms, with focus on CPU usage, memory usage, display rendering, and more. Apple claims apps can launch 30% faster on the iPhone and iPad, while new photos will show up in the library up to 70% faster. The advanced CPU scheduler on the iPhone is being optimized and brought all the way back to the iPhone 11, which will be the oldest iPhone supporting iOS 27.

Network transitions — especially between cellular and Wi-Fi — were also highlighted, so you don't have to toggle as often.

The company also highlighted a new content index across Spotlight, photos, mail and more, which will index your existing files immediately and continue to do so as new content comes in. Apple demonstrated a new ranking system for mail on Mac, showing more relevant results, even if it's years old.

AI and Siri

Federighi claimed that the mission is to turn AI into helpful and intuitive products, rather than AI for the sake of AI.

"Truly helpful AI must be centered around you and your needs," Federighi said, saying it needs to be designed for devices, apps, and with privacy in mind. Surprisingly, Federighi briefly mentioned that its foundation models are integrated with Gemini and will be used "on-device and in servers" with private cloud compute. Additionally, Apple has a more powerful on-device model.

Apple SVP for Siri engineering Mike Rockwell showed off a revamped Siri. The new Siri grows out of the dynamic island with "Hey Siri" or when you hold the buttons. It retains context between interactions.

A new voice experience includes a more conversational tone. Voice can now be customized for expressivity and pace, with sliders to adjust them in a new Siri app.

Rockwell asked about a Suki Waterhouse concert, and Siri told him when the concert will be. You can even inquire about tickets, which in this case were part of a lottery. Rockwell asked to be added to the lottery and to play her newest single in separate requests, as it remembered context.

He also demonstrated finding photos from a recent trip, asking Siri to pick out photos that include just specific people, and then share them with family members via text – without even having to go into the Photos app.

On iOS, you can swipe down, use the side button, or say "Hey Siri." On the Mac, Siri is integrated into Spotlight, or you can use Siri requests in system menus. One demonstration included using Siri on a Mac to help build a Maker Space at a child's school. When "How should I think about building a maker space in a shed” was typed into Spotlight, macOS realized it was a question specifically for the assistant. Siri was able to search through texts from the demonstrator's son, to address lingering electrical issues and find a fix.

Siri AI is also coming to iPadOS, where you can use the dedicated Siri App. You'll be able to see a conversational history synced with iCloud across devices, including Mac, Apple Watch, and iPhone. On visionOS, you can simply look at Siri and initiate your query.

Visual Intelligence will be integrated into the iPhone camera app. You can tap the shutter button for Siri AI to “see what you see” and provide contextual information, powered by Apple Foundation models. Pointing your iPhone at a plate of food can give you nutritional information, while showing the camera an image of a restaurant bill will let you split it among friends. You can also use visual intelligence to ask about items presented on your screen.

You'll also be able to write with Siri anywhere you type. Siri can generate drafts, including emails, and will provide suggestions and automatic proofreading, available systemwide, even in third-party apps.

Siri AI won't be available in the European Union and China initially, as Apple deals with international regulation and privacy laws. Siri AI will first be available in English.

Apple Intelligence in apps

Apple also demonstrated how Intelligence can be used within and across applications, including Safari, Passwords, Photos, and Home.

In Safari, Apple has introduced tab management features, which can organize tabs into topics. Another option lets you monitor a page and asks you to "Notify me" for changes on a page, like a product coming back in stock or a ticket becoming available. You can then close the tab, but Siri will notify you when the page has changed. Perhaps the most impressive change is one to Shortcuts. You can describe what you want a Shortcut to do in natural language, and it brings together the steps automatically. One example: “When I’m leaving home, message Pedro” brings together a shortcut using Maps and Messages, which should open up Shortcuts to a much wider variety of people who may have been put off by its complicated nature.

Other demos included custom Safari extensions to adapt web pages for you, a Password app that can automatically fix compromised passwords for you on "eligible" accounts. In messages, you can get one-tap suggestions, and the phone app can find flight information when calling an airline.

For Home, Apple Intelligence can understand how a number of unrelated notifications work together to create fewer notifications and provide images from multiple security cameras. It can even track package deliveries across multiple cameras. 4K resolution will work on supported cameras.

In the Image Playground, you can create high-quality images in any style, with the image generation model running on private cloud compute. You can use natural language to adjust existing images and use it across the device, such as on your lock screen. The Photos app may have the most controversial aspects, as you can extend photos to change photo borders without cropping. There's also an option to reframe the entire scene with a spatial camera, as if you were moving the lens in the original shot. This feature works on any image in your library, even if you've taken the photos on other cameras or phones.

These features will support all languages Apple Intelligence works with. Some features will have usage limits that will adjust based on iCloud+ subscriptions.

Developers will have access to improved models, including server models, through a new API that will support natural language and images.

Lastly, Apple highlighted updates to XCode, including agentic coding, choosing the model and agent of your choice (including the newly added Google Gemini). It can connect to Figma and GitHub for design, and a new Device Hub will let developers simulate devices, including touch screens, alongside real-world testing devices.

Trust and Safety

Apple also demonstrated new child safety tools, with new child accounts and parental controls. Child accounts will automatically block adult websites and implement app store age restrictions.

Children will be able to ask parents for permission to buy or download apps in messages, as well as browse new websites in Safari for children under 13. There will be similar permissions for contacting new people outside of your family.

A new safety communication warns about nude images and intervenes, but will also be expanded to gore or violent content. Parents will have Time Allowances across entertainment, games, and social media, with shared allowances across all three, or you can set them individually. A redesigned Screen Time will let parents see how devices are being used. Developer tools are being made more capable for third-party apps to take advantage of the new features.

Golden Gate will be the first version of macOS to exclusively support Apple Silicon Macs. Last year, Apple announced that Tahoe would be the final major release to support Intel-based systems. (Intel-based Macs will still receive three years of security updates.) This is the last macOS release to support Rosetta 2, the tool that lets Apple Silicon computers run Intel applications through an emulation layer (though parts may stick around to keep legacy games running). Without Intel processors to support, Apple is urging developers to make native applications for its own silicon.

Support and Release Dates

Developer betas for the 27-suite of OS releases are available today, with public betas starting in July. Final releases are expected in the fall.

macOS 27 Golden Gate will be supported on devices with Apple Silicon. Apple's website specifies:

• MacBook Neo
• MacBook Air (2020 and later with Apple Silicon)
• MacBook Pro (2020 and later with Apple Silicon)
• iMac (2021 and later with Apple Silicon)
• Mac mini (2020 and later with Apple Silicon)
• Mac Studio (2022 and later with Apple Silicon)
• Mac Pro with Apple Silicon (2023)

However, adjustments to Siri's voice and advanced dictation will require newer chips on the Mac, iPad and iPhone. On the Mac, that means "an M3 and later and at least 12GB of unified memory." Otherwise, you need an iPhone 17 Pro, iPhone 17 Pro Max, iPhone Air, iPad models with M4 and later and at least 12GB of unified memory, or Apple Vision Pro with M5.

Thanks to AI-assisted security research, hackers with hats of various colors are finding exploits everywhere. Linux has had its worst week in years with the CopyFail and Dirty Frag root-gaining vulnerabilities, and things aren't much rosier at Microsoft, thanks to the YellowKey BitLocker bypass, as well as GreenPlasma and RedSun privilege-gaining exploits. Now, it's Apple's turn with a local privilege escalation that gets past the M5 chips' much-vaunted Memory Integrity Enforcement (MIE).

There aren't many technical details, but the vulnerability is simple in practice: run a command as a standard user and gain root (administrator) access to the machine. Macs are rarely servers, so the practical impact is limited. However, the exploit remains concerning, as it's relatively easy to trick a user into running it and, with full system control, also hard to find and remove. The research team in question is named Calif, and as far as they know, the boffins there are the only ones making a public disclosure of this issue. Such assumptions are tricky in this day and age, though.

Mercifully for Captain Cook's ship, instead of being a zero-day reveal out of nowhere that left systems administrators scrambling, the exploit in question was disclosed to the company in advance (in person, no less). Calif published the vulnerability overview as part of a series of blog posts called the Month of AI-Discovered Bugs, since this new Apple vulnerability falls within a set of security findings aided by AI tools — in this case, Anthropic's Mythos Preview.

The researchers tested their code on an Apple M5 machine and macOS 26.4.1. The exploit chain impressively sneaks past MIE, a security feature present on M5 and A19 chips that labels each 16-byte memory slice with a 4-bit tag associated with the pointers that use it. MIE is enforced at the hardware level in a hypervisor-like configuration and effectively protects against most common classes of security exploits, namely, but not only, buffer overflows and use-after-free vulnerabilities.

As an oversimplification, MIE ensures that any memory read or write operation acts on the data that it was originally meant to, even at the kernel level. If that doesn't happen, either your application has a bug, or someone's up to shenanigans. The base feature is part of ARM MTE, and MIE is an Apple-added layer that enforces the said checks at the hardware level, with purportedly little to no performance overhead, and only 3% memory wastage. This blog post goes into more detail on the subject, and it's quite an interesting read.

Sources close to both Apple and Intel have said that the two companies have reached a preliminary agreement for Intel to manufacture processors for Apple. According to The Wall Street Journal, the two parties have been in talks for over a year, and they’ve been finalizing a formal deal over the past few months. Apple is reportedly looking for alternative fabs to TSMC, as it wants to diversify its supply chain in the midst of ongoing chip shortages, and it seems that this latest development is a confirmation of this report.

It’s currently unclear which chips the U.S. semiconductor company will be making for the largest consumer electronics firm in the world, but it previously made the x86 processors used in Macs and MacBooks from 2006 to 2023. Intel also had the chance to build the A-series chips that Apple used for the iPhone and iPad, but it fumbled the opportunity — with Tim Cook complaining to TSMC founder Morris Chang, “Intel just does not know how to be a foundry.”

Intel apparently learned its lesson, though, as Apple was reportedly considering Intel’s 18A process for its entry-level M-series chips as early as November of last year. Even before this, the two companies were already in discussions about a potential investment into the chip maker, something that President Donald Trump alluded to in early 2026. Aside from Apple, Intel has also received a $5-billion investment from Nvidia and that the two are partnering to develop an x86 RTX SoC for PCs. Elon Musk also tapped Intel for his TeraFab project, which will use the company’s 14A process to make AI chips.

These developments are excellent news for Intel CEO Lip-Bu Tan, who has been working hard to put the company back on track after former CEO Pat Gelsinger announced disastrous results in July 2024. It has also caused the company’s stock price to skyrocket, hitting a record high of $126.23 at the time of writing and beating its former peak during the dot-com boom of 2000. But we'll have to wait for confirmation from the two companies — as neither has officially commented on this news yet — to learn the details of this groundbreaking deal.

Apple, arguably the largest phone and computer manufacturer in the world, isn’t immune to the chip shortage, as it once again reduced the highest memory capacity available for the Mac Studio and Mac Mini. Apple's Mac Studio could be configured with up to 512GB of Unified Memory at launch, which is more than enough for professional usage and local AI applications.

Now, if you visit the Apple Store, you can only purchase the Mac Studio with a maximum of 96GB of Unified Memory, while the Mac mini is only available with up to 48GB (down from 64GB).

Apple's Unified Memory architecture has become popular for running local AI models and agentic AI applications like OpenClaw. When the latter exploded in popularity in early 2026, many AI developers and enthusiasts jumped at the opportunity to acquire a small PC with a very high memory pool. This resulted in a shortage that pushed delivery times from six days to six weeks.

This situation, combined with the global memory chip shortage, has forced Apple to discontinue the 512GB SKU and increase the price of the 256GB upgrade to $2,000 in early March. However, the increase in price did not dissuade buyers, as company chief Tim Cook warned during the company’s latest earnings call that supplies for both the Mac Studio and Mac mini could remain constrained for months.

Despite the increased prices and the removal of higher-capacity versions, delivery times for both the Mac Studio and Mac Mini haven’t improved. The M4 Max Mac Studio with 64GB of Unified Memory can take six to seven weeks to deliver, while all other configurations, including the 36GB and 96GB variants, can take between nine and 10 weeks. Prospective buyers of the M4 Mac mini with 16GB of Unified Memory may fare slightly better, with a three-to-four-week delivery window; this worsens to 10 to 12 weeks if you upgrade the memory to 24GB or 48GB.

We’re also starting to see shortages affect high-capacity MacBooks. While 24GB and lower-capacity MacBook Air laptops are still thankfully readily available, you need to wait about two to three weeks if you want the 32GB version. 48GB versions and up of the MacBook Pro also now entail a wait of one to two weeks, while even the budget MacBook Neo, which became insanely popular among students and entry-level buyers, takes around two to three weeks to deliver.

Cook said during the company’s first earnings call of 2026 that the company is chasing memory supply and that it will have a greater impact on its Q2 earnings. We’re now feeling the brunt of the shortage, as we see Macs, high-capacity MacBooks, and the entry-level MacBook Neo experience extended delivery times. Thankfully, mid-range buyers can still easily acquire the standard iMac, entry-level MacBook Airs, and MacBook Pros.

Apple is reportedly talking with Intel and Samsung to produce processors for the company as it’s grappling with a shortage of chips for its latest products. According to Bloomberg, Cupertino has had multiple early-stage discussions with Team Blue, while some key Apple executives have also visited a Samsung plant that’s still in development — likely its Taylor, Texas, fab, which is set to start risk production this year.

None of the talks have resulted in any orders, so far, the report states. There have been concerns about using non-TSMC tech in Apple products, especially as it has been producing chips for the iPhone since the A8 used in the iPhone 6 and iPhone 6 Plus. However, the AI infrastructure build-out is negatively affecting the company, with Apple CEO Tim Cook conceding that it’s constrained by TSMC’s supply of advanced chips. Nvidia CEO Jensen Huang said that his company has dethroned the consumer electronics giant as TSMC’s number one customer, and even alluded that Apple might have to pay more for its chips. But even before this, Cupertino has reportedly been considering using Intel’s 18A process to build future M-series chips, especially as global geopolitical events are threatening the stability of the global supply chain.

If we look at Apple’s history, both Samsung and Intel have closely worked with the company for years. The first iPhone, all the way to the iPhone 5S, was powered by Samsung chips, while Macs and MacBooks used x86 Intel processors from 2006 until they were replaced by Apple silicon in 2023. If successful, these talks would resume Apple’s partnership with one (or both) of the companies.

This would be a win for Intel as it’s looking to land a major customer for its foundry business — something that it fumbled in 2011 when Cupertino first approached it for its custom chip needs. A move like this could also help bolster Samsung’s foundry business, as it sits at a distant second place when compared to TSMC. At the same time, it would reduce Apple’s risk of relying on a single supplier for the majority of its advanced chip supply. Even though TSMC’s Arizona plant is ramping up production and it’s estimated that it will deliver 100 million chips for Cupertino this year, it is just a fraction of Apple’s demands, with the majority of its advanced SoCs still expected to come from Taiwan.

Tim Cook has been well aware of the situation with TSMC for years now. He even said in a 2022 all-hands company meeting, “Regardless of what you may feel or think, 60% coming out of anywhere is probably not a strategic position.” This was likely highlighted with the AI-driven chip shortage. “The primary constraint is the availability of the advanced nodes our SoCs are produced on, not memory,” the Apple CEO said during the company’s latest earnings call. “I believe it will take several months to reach supply-demand balance.”

But even if Apple were to strike a deal with both Intel and Samsung today, it will take some time for production to ramp up, and consumers won’t feel its effect for several months, if not a couple of years.

Many think that Apple's history creating computers with graphical interfaces starts with the venerable Macintosh, but the company's first foray into designing GUIs actually started with 1983's Lisa. The machine was the first to bring a popular window application system and mouse control to general availability at an edible (if bitter) price of $9,995 dollars of 1983 vintage, equivalent to nearly $34,000 today.

Thanks to enthusiast and Youtuber Alex Anderson-McLeod, you'll soon be able to roll your own much-improved LisaFPGA clone. Alex designed a roomy one-board system centered around an Artix 7-100T FPGA as the Lisa's brains, coupled to 2MB of SRAM and emulated hard drive, floppy, serial, and naturally, keyboard and mouse connectors.

The clone is far superior to the original, thanks to its reliance on modern hardware and Alex's design smarts. For starters, you don't need a video converter, as the board natively outputs video HDMI with a scanline option; both its main video modes are switchable on-the-fly. Although the board includes original-spec connectors for input devices, it also helpfully has USB ports for keyboard and mouse so you don't have to procure original versions.

In a similar fashion, the serial ports can be redirected to the main USB-C connector thanks to the inclusion of a USB hub, saving you having to find USB-to-serial adapters and chunky DB25 connectors. Floppy images can be loaded from an SD card, direct connection to PC, or an original floppy drive connected to the the corresponding port.

Even for its time, the Lisa was pretty slow, thanks to its software design choices and and use of the 5 MHz variant of the Motorola 68000 CPU. As a nod to these limitations, Alex included two overclocking multipliers that can take the machine to the equivalent of 75 MHz, likewise selectable on the fly with physical switches.

The revision of the LisaFPGA featured in the video is version 2, but Alex says that version 3 is coming soon with a handful of fixes. He says the project will be fully open-sourced on Github as soon as he gets his proverbial ducks in a row. He's also considering selling the clones, and will be speaking about them at Vintage Computer Festival Southwest later in the month.

Although it was generally considered a failure due to its price and software availability, the Lisa paved the way for 1984's Macintosh, the $2,459 machine that popularized graphical interfaces and the Apple brand worldwide. The Lisa computer had an undignified ending, too, with 2,700 of them dumped in Logan, Utah, as part of a tax write-off on unsold inventory. Still, similar to how the first iPhone had major issues in hindsight, the Lisa broke ground for much greater things.

Although Apple popularized graphical home computers for the masses, Apple did not invent the GUI, nor the mouse. The GUI was Xerox's doing at its PARC laboratories, that Steve Jobs visited and, cough, cribbed notes from. The Xerox Star 8010 workstation was the first computer with both technologies, but it was sold as part of a larger package and commanded the steep price of $16,595 in 1981, or around $60,000 today. As for the mouse, it was designed at the Stanford Research institute much earlier in 1963.

Speaking during Apple’s second fiscal quarter 2026 earnings call, CEO Tim Cook warned that supplies of the company’s Mac mini and Mac Studio desktops could remain constrained for months. According to Cook, surging demand driven by artificial intelligence workloads outpaces Apple’s manufacturing capacity.

The surge comes amid growing interest in “local AI,” where models run directly on personal machines rather than on remote cloud servers. Privacy concerns, latency reductions, and rising cloud inference costs have pushed many developers and companies toward on-device AI processing.

Cook’s warning marks one of the clearest signs yet that the AI boom is beginning to reshape the personal computer market in ways that extend beyond traditional GPU manufacturers like Nvidia. The comments also officially confirm our report that some Mac models were facing significant shortages and shipping delays, primarily driven by an "ordering frenzy" for high-memory configurations, spurred by demand to run local AI agents, such as the "OpenClaw" mode.

Apple’s Mac mini and Mac Studio have become particularly attractive for AI development because of the company’s Apple Silicon architecture. Unlike conventional desktop systems that separate CPU and GPU memory, Apple’s unified memory architecture allows AI models to access large pools of shared high-bandwidth memory more efficiently. This makes the systems especially useful for running local large language models, AI agents, and inference workloads.

The higher-end Mac Studio configurations can also be equipped with massive amounts of unified memory, allowing developers to run increasingly large AI models directly on desktop hardware. Combined with relatively low power consumption, the systems have gained popularity among AI developers seeking alternatives to expensive server-grade hardware.

Industry observers have noted rising interest in Apple desktops from AI enthusiasts over the past year. Online developer communities have increasingly discussed using Mac Studio systems for running open-source AI models locally, particularly as demand for high-end AI GPUs continues to strain global supply chains.

Apple’s supply warning also arrives during broader pressure across the semiconductor industry. Advanced chip packaging technologies and high-bandwidth memory production have already been strained by soaring demand for AI infrastructure. Several semiconductor firms have warned of prolonged shortages tied to AI-related manufacturing bottlenecks.

The constraints may also reflect Apple’s growing ambitions in artificial intelligence. The company has been steadily expanding its AI strategy following the introduction of Apple Intelligence across its ecosystem. While Apple has historically emphasized on-device AI processing for privacy and efficiency reasons, the recent surge in demand for its desktop systems could further strengthen its position in the emerging local AI computing market.

The shortages could persist for “several months,” according to Cook, suggesting customers may face extended shipping delays for some Mac mini and Mac Studio configurations.

After failing to reignite interest in the Vision Pro with the release of the M5 model, Macrumors reports that Apple may have just about given up on the product. Initially launched on February 2, 2024, the device failed to garner the popularity typically seen with Apple products, reportedly selling only around 600,000 units. For context, the company sells hundreds of millions of iPhones each year.

At $3,499 and 1.3 pounds, the device's price and weight have been cited as contributing factors to low sales. There were initial reports dating back to 2024 that Apple was winding down production of the Vision Pro. However, the company released an updated version in late October 2025, featuring a faster M5 chip. The chip resulted in a higher 120Hz refresh rate, 10% more rendered pixels, and around.

The model also came with a redesigned headband with an upper strap to better distribute the headset's weight. Despite the additional features, the price remained $3,499. Yet, according to MacRumors, sales did not improve. Apple reportedly received an unusually high percentage of returns, far exceeding those of any other modern Apple product.

According to Bloomberg journalist and Apple news Mark Gurman, Apple had broken up the Vision Products Group in 2025, splitting it between software and hardware engineering. In an X post yesterday (expand the tweet below), April 29, he claimed that Apple had reassigned much of the Vision Pro software team to Siri, and the hardware team to smart glasses, citing CEO John Ternus’ stance against the Vision Pro as a product.

However, this claim is being met with widespread skepticism, as readers were quick to point out what they believed were contradictory job board posts indicating that Apple was actively recruiting for the Vision Production Group. We can confirm that Apple has indeed posted several of these roles on its official job board over the past few months.

On the other hand, the job posts do not explicitly confirm the continued development of the Vision Pro. One of the posts read as follows: “Originally developed to help VisionOS achieve its critical motion-to-photon latency targets—providing users with a seamless, realistic experience through Vision Pro—our technology is now expanding to iOS and macOS.”

The mixed reports surrounding the Vision Pro have left consumers wondering exactly what is going on, especially as there has been no official report. What’s highly plausible is that the company may be developing smart glasses, a lighter, more streamlined AR offering that is still years away from consumer release.