M.2 2230 drives are a bit of a luxury, packing in high levels of performance in a small space. These drives always carried a premium, which makes for a challenging purchase decision, and the reasons for the premium have changed. Before, the premium was due to this form factor being a niche segment with few options, and now it’s due to a memory squeeze that makes high-capacity drives much more expensive. So when a drive like the PNY CS2342 comes along, which is available in 2TB at an accessible price, it’s worth paying attention.

The CS2342 has known, mature hardware that provides excellent performance without producing a lot of heat. This is important for this form factor. While it’s far from the most efficient drive we’ve tested, it’s good enough. You can get good flash with a 2TB drive in a single-sided form factor, which, with reasonable power efficiency, is all you need. There are drives that will perform better in some scenarios, there are more efficient drives, and there are drives that could be less expensive with QLC flash. However, if the goal is simply to acquire a competent, general-purpose drive, this model is an excellent choice.

PNY CS2342 Specifications

The PNY CS2342 is available at 1TB and 2TB, priced at $269.99 for the latter at the time of review. The 1TB was out of stock, but there are other options available if you’re willing to compromise on speed or flash type, like the Kingston NV3 (2230). The CS2324’s pricing at 2TB is good for what it is – a high-speed PCIe 4.0 SSD with TLC flash, in this form factor – but you can save some money by going with QLC. The Crucial P310 is still in stock for the intrepid.

If you are instead looking for an M.2 2242 solution – and you can extend an M.2 2230 drive if that makes more sense – then your options right now are also limited. The Corsair MP600 Micro, which is the M.2 2242 version of the MP600 Mini (E27T), is priced at $189.99+ on Amazon and is not a bad choice. Alternatively, the Transcend MTE410S, which we have not yet reviewed, is available at $183.90, and we would recommend this drive if you are primarily concerned with reliability or have a PCIe 3.0 device. It uses an SMI SM2269XT – see our Lexar Play (2230) review to get a feel for this controller – but the MTE410S uses older BiCS5 flash in comparison to other options. The CS2342 could still work here with an extender as a compromise.

For performance, the CS2342 can hit up to 7,300 / 6,000 MB/s for sequential reads and writes, with no IOPS specs given by the manufacturer, although we know this hardware can hit around 1 million IOPS. If you are interested in the latter, you can find details in the ATTO & CrystalDiskMark benchmarks on the following page. The drive has a standard five-year warranty with up to 600TB of writes per TB of capacity.

PNY CS2342 Software and Accessories

PNY’s website has two main downloads for its SSDs: an SSD Toolbox and Acronis software. The former is also touted as a firmware updater. SSDs toolboxes are applications that are designed to monitor and diagnose issues, support features for performance, which include testing, and act as a central point for storage management. These toolboxes can let you access secure erase and encryption functions on supported drives, but the firmware updating part is probably the most important. While some also include data management, Acronis provides OEM versions of its True Image software to handle cloning and backup when preparing your new drive. PNY covers both sides, which is nice if you prefer not to mess with random programs.

PNY CS2342: A Closer Look

The PNY CS2342 is a single-sided drive in the M.2 2230 form factor. This means it has a DRAM-less controller and a single NAND flash package. Despite using TLC flash, which is faster with higher endurance than QLC but often means lower capacities, the drive can reach 2TB with no problems. This makes it good if you’re looking for the best of everything in this form factor. The drive’s label lists power draw at 3.3V/2.6A, which is below 9W, but the drive is rated at 5.70W peak via SMART and, in practice, will pull less than this.

The drive has three primary components: the SSD controller, the NAND flash package, and the power management IC. The main downside to having an M.2 2230 SSD is that all of this componentry is smashed together, which can worsen issues with heat. This is especially true with newer drives because they are extremely fast. Luckily, most of the time the drive won’t be at full power, and this amount of heat can be handled with basic solutions. If your host device lacks cooling for the drive, we recommend going for thermal padding or a do-it-yourself solution.

We do not recommend throwing a slab of metal on it – many cheap M.2 2230 SSD heatsinks are a flat piece of metal – as while that works for heat spreading, it’s far more effective on M.2 2280 drives where you actually have distance between the components or more total component surface area. Transferring the heat away from the drive with thermal padding is a better solution. Some host devices may have enough headroom for low-profile heatsinks, which are even better.

MORE: Best SSDs

MORE: Best External SSDs

MORE: Best SSD for the Steam Deck

Comparison Products

We have a small array of drives going up against the PNY CS2342 today. You can get a good feel of what drive is right for you, even with a small list. We first have the Crucial P310 (2230), which is surprisingly fast for a QLC flash drive, but its availability may become limited once the current supply runs out, as Micron is pivoting away from consumer memory products. The Corsair MP600 Mini is basically the same as the PNY CS2342, but pricing and availability vary.

Turning to older drives, the WD Black SN770M was a performance champ and can still win out in some tests. It’s very popular, but pulls more power. The other TLC drives in this range are like the Sabrent Rocket 2230, which basically uses a slower Phison controller but with comparable flash. This doesn’t matter as much for the PCIe 3.0 Steam Dec,k which is speed-limited. You can also go to QLC flash to save money on the higher capacity options with either the Silicon Power UD90 (2230), with the Rocket’s Phison controller, or the Kingston NV3 (2230) with an SMI controller that technically competes with the P310’s.

Trace Testing — 3DMark Storage Benchmark

Built for gamers, 3DMark’s Storage Benchmark focuses on real-world gaming performance. Each round in this benchmark stresses storage based on gaming activities including loading games, saving progress, installing game files, and recording gameplay video streams. Future gaming benchmarks will be DirectStorage-inclusive and we also include notes about which drives may be future-proofed.

We like what we see here. The CS2342 performs very well, only falling behind the QLC-based P310. While the P310’s hardware and optimization enable it to edge out the competition here, under poor conditions – if the drive is fuller after a large update – the CS2342 with TLC flash would probably be more consistent. This is because performance will drop more for a QLC drive as the flash is slower when doing operations outside the cache, which includes necessary internal maintenance tasks. We think any of these drives down to the Black SN770M would give the best experience.

Trace Testing — PCMark 10 Storage Benchmark

PCMark 10 is a trace-based benchmark that uses a wide-ranging set of real-world traces from popular applications and everyday tasks to measure the performance of storage devices. The results are particularly useful when analyzing drives for their use as primary/boot storage devices and in work environments.

The drive is a bit slower in PCMark 10. It’s in good company, though, getting close to the MP600 Mini (E27T) and Black SN770M, which are performance drives. We wouldn’t expect the NV3 to be faster in edge cases because it uses QLC flash, as defined above. The P310 does run away with the lead in this one, though, and since application performance will often be read-heavy – you may want responsiveness in a wide range of apps simultaneously, too – it probably provides the best experience even in less ideal circumstances.

Steam Deck Testing — Gaming, KDiskMark, and Temperature

The Steam Deck is not the only portable gaming system in town, but it was the first and most popular to take M.2 2230 SSDs. While some systems have moved on to fit 2280-length drives, 2230 and 2242 remain popular for many systems, and such drives will work fine in longer slots with the proper standoff or extender. The Deck operates in PCIe 3.0 mode for its SSD, which limits maximum bandwidth but has less of an impact on responsiveness. The Deck is also still useful for gauging drive temperature and power efficiency. Drives scale up/down with link speed to a predictable degree, and our Windows tests also provide insight into PCIe 4.0+ performance.

Our current testing for the Deck involves analyzing game load times for some popular games. This is probably the most important metric for gamers, but the difference between one SSD and another can be small. We also use KDiskMark, a CrystalDiskMark-like substitute that uses the flexible I/O (FIO) tester instead of Microsoft’s diskspd for its benchmarks. We also check the drive’s maximum temperature during this test to make sure that no throttling occurs.

The tests in this section are run under the stock Arch-based SteamOS Linux platform, but our other tests are conducted as per our normal reviews using Windows. Many portable gaming systems today use or can use Windows solo or in a multi-boot solution. This testing section is designed to give an idea of performance under Linux with the Proton compatibility layer.

Does a second or two matter with loading games? Not really, although it can add up, and it’s easy to become impatient with multiplayer games in particular. It’s still worth knowing how fast games load on a drive because this scales to some degree. When the drive is very full and stressed with big updates, you want to know you can still load without lag. We think the CS2342 is great here if you want the best Deck or mobile experience.

Transfer Rates — DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with a custom, 50GB dataset. We write 31,227 files of various types, such as pictures, PDFs, and videos to the test drive, then make a copy of that data to a new folder, and follow up with a reading test of a newly-written 6.5GB zip file. This is a real-world type of workload that fits into the cache of most drives. It also has the highest IOPS for QD1 random reads, which is the number one metric people look at for real-world feel. We’re only disappointed by the QD1 sequential read performance, which is, thankfully, still good enough. The drive manages this while staying very cool, which reduces the likelihood of throttling, which is very important in the Deck and other host devices.

You won’t be copying at speed a lot in a Steam Deck, but other devices can benefit from faster transfer performance. The CS2342 lands in a good spot. It has ample performance for tasks on handheld devices and has a significant lead over old favorites like the Black SN770M.

Synthetic Testing — ATTO / CrystalDiskMark

ATTO and CrystalDiskMark (CDM) are free and easy-to-use storage benchmarking tools that SSD vendors commonly use to assign performance specifications to their products. Both of these tools give us insight into how each device handles different file sizes and at different queue depths for both sequential and random workloads.

What we see are nice, clean ATTO curves from the CS2342. CDM sequential performance is also good, although the SN770M is a standout with QD1 reads. This metric is a good indicator of game and app loading performance, as many do have sequential workloads. Random 4K QD1 reads usually have more focus, and here the CS3242 finishes first by a whisker. This is excellent performanc,e and the drive will feel incredibly responsive. Random write latency is also fantastic, so this is a drive that can do more than just game.

Sustained Write Performance and Cache Recovery

Official write specifications are only part of the performance picture. Most SSDs implement a write cache, which is a fast area of pseudo-SLC (single-bit) programmed flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the cache and into the "native" TLC (three-bit) or QLC (four-bit) flash. Performance can suffer even more if the drive is forced to fold, which is the process of migrating data out of the cache in order to free up space for further incoming data.

We use Iometer to hammer the SSD with sequential writes for 15 minutes to measure both the size of the write cache and performance after the cache is saturated. We also monitor cache recovery via multiple idle rounds. This process shows the performance of the drive in various states as well as the steady state write performance.

The PNY CS2342’s first mode is a very fast but temporary pSLC write mode where it writes at 6.06 GB/s for over eight seconds. The cache is therefore around 52GB, which matches our expectation for this hardware, which usually has one around 50GB. This is a very small cache with resultant trade-offs. It can’t cache as many writes, but its performance is much more consistent. Personally, we feel this usually provides the best experience as smaller, random, or bursty writes can still fit in the cache, which gets smaller as the drive fills, but remains adequate in size and performance due to not needing a lot of resizing, and your extreme edge case scenarios don’t feel as bad.

We’ve suspected in the past that Phison, the controller manufacturer, did this intentionally after getting negative feedback on some drives in the past. This is interesting for a drive in this form factor, as often you see fewer writes, but it’s great for enclosures where you’ll be seeing more writes. M.2 2230 drives often see a lot of movement as upgrade drives or in small enclosures, though, so it works. The change also accommodates our criticisms while fitting into our 50GB tests. We’re glad Phison took the feedback seriously and looked at real-world scenarios when going back to the drawing board.

Post-cache performance in the TLC mode is at around 800 MB/s. This is a good result and would be great for a 10Gbps enclosure if that’s your plan. In fact, this would be extremely effective for that type of usage, especially if you want a small form factor enclosure.

Power Consumption and Temperature

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is an important aspect to consider, especially if you're looking for a laptop upgrade, as even the best ultrabooks can have mediocre stock storage. Desktops may be more performance-oriented with less support for power-saving features, so we show the worst-case.

Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt, or efficiency, is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state more quickly, ultimately saving energy.

For temperature recording, we currently poll the drive’s primary composite sensor during testing with a ~22°C ambient. Our testing is rigorous enough to heat the drive to a realistic ceiling temperature.

The CS2342 is efficient but on the lower end compared to newer drives. If power efficiency matters to you, this might put it lower on the list. It is still efficient enough for its intended usage, and we don’t see any problems there. The Black SN770M is more than 20% less efficient, which is substantial, but even that didn’t have heat issues in our testing, or really hurt battery life in a significant way. We do think the Black SN770M will run hotter in many cases, so the CS2342 is a good option if you want good performance but less throttle potential. Both drives top out at under 5W, which should be manageable in most cases.

Test Bench and Testing Notes

We use an Alder Lake platform with most background applications such as indexing, Windows updates, and anti-virus disabled in the OS to reduce run-to-run variability. Each SSD is prefilled to 50% capacity and tested as a secondary device. Unless noted, we use active cooling for all SSDs.

PNY CS2342 Bottom Line

The PNY CS2342 excels in all of the important areas for a drive in the M.2 2230 form factor. 2TB of space, strong real-world performance, reasonably good power efficiency, and it’s single-sided, too. This is great, although if you can use a 2242 drive, then 4TB is a reachable goal with the Corsair MP700 Micro. Samsung also has an OEM drive out for 4TB in that form factor, a shorter model of the PM9E1, but that is double-sided. Its advantage lies in the presence of DRAM and maximum PCIe 5.0 performance, and your other option is probably some variant of SK hynix’s BC711. Systems limited to 2230, like the Steam Deck, have to be more picky, but the CS2342 is perfect for it.

We’ve listed the drive’s power efficiency as a negative because it’s simply average by this point in time. This is not a big drawback in most cases, as it achieves a good power-performance balance. The impact on battery life won’t be large, and it’s efficient enough to avoid throttling, but some systems or projects may be more needy than others. There are times when you might want a different drive for performance in specific areas – the Crucial P310 (223) is surprisingly fast for a QLC-based SSD – or want that extra power efficiency, but overall, the PNY CS2342 will provide a good all-around experience. At the time of review, you can find the 2TB model for a reasonable price, and we strongly recommend it.

MORE: Best SSDs

MORE: Best External SSDs

MORE: Best SSD for the Steam Deck

11/11/25 update: Representatives from PNY reached out to us clarifying that, while the company's plans have indeed shifted due to NAND supply issues, there will, be new deals coming from the company this year, and not all of the previous deals may be cancelled or changed. We have removed text in the headline and body of the story that previously stated that PNY will have no new deals.

As the AI-driven shortage of NAND tech continues to bite and enthusiasts chart 100% price increases on RAM for their builds, popular PC hardware vendor PNY has confirmed that it is suspending its upcoming Black Friday and Cyber Monday promotions on flash USB storage, citing rising NAND costs.

PNY had previously announced a series of deals rolling out across November, with early November savings of up to 60% on some of its top USB products. For instance, its Elite Turbo Attache 4 64GB USB 3.2 Flash Drive is currently just $7.99 at Best Buy instead of $19.99. All of the discounted products planned for November were USB Flash drives of one description or another, ranging from cheap low low-capacity options to up to 1TB of 600MB/s storage.

Tom's Hardware was contacted by a third-party PR representative of PNY on November 10, who advised us that the deals shared on behalf of the company were no longer valid. Crucially, we were also told that no new deals are being put in place on these products. Some of the early discounts of up to 60% are still being offered on Best Buy's website, so this seems to indicate that the further discounts on these products planned for the Black Friday/Cyber Monday weekend are no longer on the cards. Some of those planned reductions were discounts as high as 70%. While some remain on sale at previous discounts, others have returned to MSRP at Best Buy. PNY was not immediately available when approached directly for comment.

The sweeping suspension of an entire brand's worth of USB storage deals could be a bellwether of tough things to come for consumers. Notably, if NAND shortages are trickling all the way down to flash drives, the impending market crunch could be worse than first thought. While Early Black Friday deals aplenty can be found on a new gaming chair, peripherals, monitors, and components like some GPUs and CPUs, the best SSD deals and memory savings are a little softer. As per our previous coverage, enthusiasts trying to build PCs on popular hardware forums are charting price increases on DDR5 kits of 100% or more. In one notable case, one beleaguered builder cited that the 64GB Kingston FURY Beast kit they'd marked out for their build is now more expensive than the Ryzen 7 9700X CPU it was being paired with.

It is impossible to say which immediate factor may have driven PNY's decision to suspend its Black Friday sales. It could be that the company is worried it doesn't have enough stock to sustain it through the holiday season, or it could be a preemptive move to ensure flash drive stock remains heading into the New Year. It is also unclear whether there will be any knock-on effect on PNY's supply of consumer SSDs, including its new CS3250. We continue to await comment from PNY and will update this story accordingly.

I'm constantly tracking the Nintendo Switch 2, and it's must-have accessories and storage options since their release. Stocks of anything Switch 2 have been flying off retailer shelves faster than you can shake your fist at the sky in frustration. The Switch 2 has proved to be an insanely popular games console, and there's currently no end in sight for when the console and microSD cards will be readily available at the correct price point.

Today, PNY's 256GB microSD Express card is back in stock at Amazon with a list price of £59. This isn't a deal, as the price hasn't changed since the launch of this microSD card, but it is in stock; for how long though is the question. If you're looking for extra storage for your Switch 2 console, then this is a good microSD Express card to snap up. There are also options from SanDisk on Amazon, but the shipping lead times are one to two months, so they might as well be listed as out of stock.

The Nintendo Switch 2 makes use of the latest microSDXC Express technology, which is a far faster storage option than the previous microSD cards that have been used for earlier iterations of Nintendo's Switch console. Using PCIe Gen 3x1 NVMe tech, the latest microSD Express technology allows speeds more similar to laptop and desktop SSDs with performance up to 890MB/s read, and 750MB/s write speeds.

The native storage capacity of the Nintendo Switch 2 is 256GB, so this 256GB microSD Express card from PNY doubles the amount of software that you can install on your console. It's microSD express technology means installation, load times, and data streaming are much faster, for a better gameplay and user experience.

Check Amazon for all microSD Express cards available for the Nintendo Switch 2 games console.

Hot on the heels of Nintendo Switch 2 pre-orders and the potentially wallet-straining news that the new handheld is switching to microSD Express storage technology, PNY has unveiled its new microSD Express Flash Memory Cards.

PNY says that its new cards are capable of up to 890MB/s read speeds and up to 750MB/s write speeds, 4.4 times faster than standard UHS-I microSD cards.

It features PCIe Gen3x1 microSD technology, which should, in theory, reduce load times of your favorite Nintendo Switch 2 games and even improve performance.

The Nintendo Switch 2 comes with 256GB of onboard storage as standard, so if you want any more, you'll need to invest in a microSD Express card like this one.

PNY says its microSD Express Cards maintain backwards compatibility with USH-I and USH-II host devices at their respective speeds.

They also come with unique heat dissipation technology and built-in microSD thermal monitoring PNY claims adjusts to optimize temperature and keep things running smoothly.

PNY's microSD Express cards are also rated as magnet proof, shock proof, temperature proof, waterproof, drop proof, X-Ray proof, humidity proof, and wear-out proof. They also come with a lifetime warranty.

PNY is probably best known for its GPUs, DRAM, and general memory products — memory cards and portable flash drives — and it's branching into the SSD market. Sometimes PNY draws outside the lines, like with its pSLC LX3030, but usually the company plays it safe by selling drives with a wider appeal. The CS3150 looks to cater to the enthusiast market, with a PCIe 5.0 SSD that stands out with its dual-fan cooler and optional RGB. It has DirectStorage-optimized firmware and hardware encryption support as well, leveraging the best aspects of its Phison E26 SSD controller.

The CS3150 isn’t the fastest drive of its kind, though it ranks near the top of our charts. It's also limited to only 1TB and 2TB capacities, with the latter being difficult to find for sale at present. PNY sent us the 1TB model with RGB lighting for this review.

The CS3150 hardware allows for good all-around and sustained performance, but it retains the power consumption issues of its high-end hardware. The good news is that the heatsink and temperature-controlled fans keep the drive cool and are much quieter than some of the actively cooled fans we've tested. That and the optional, controllable, syncable RGB are the CS3150’s biggest selling points. This could be an interesting drive for some builds, even at 1TB, although the pricing makes a niche product.

PNY CS3150 Specifications

The PNY CS3150 XLR8, or just CS3150, is available in 1TB and 2TB capacities, with no 4TB variant available or planned at this time. PNY’s MSRP is $179.99 and $299.99 for the two capacities, adding another $10 on for the RGB or Epic-X model. These prices are higher than equivalent drives like Gigabyte’s Aorus Gen5 12000, the Corsair MP700 Pro, and the Teamgroup Cardea Z540, although the Z540 comes bare.

At present, the 1TB SKUs are slightly more expensive than this on Amazon, and the 2TB is missing in action. That’s a shame as the 2TB models would be far more interesting — we can't imagine many enthusiasts who would be willing to spend a lot of money on a high performance PCIe 5.0 drive would want to stick with 1TB.

The CS3150 can hit up to 12,000 / 11,000 MB/s for sequential reads and writes. No information is given for random read and write IOPS, but we would expect a maximum of around 1,400K / 1,400K — similar to other Phison E26 1TB drives with 12 GT/s NAND. The warranty is also standard for drives of this class, rated at five years with 700TB of writes per TB capacity. The CS3150 does support hardware encryption with TCG Opal 2.0.

PNY CS3150 Software and Accessories

If you happen to get the RGB or Epic-X version of the CS3150, PNY supplies its VelocityX RGB control software. This can sync with other PNY RGB products, namely GPUs and DRAM. For benchmarking and drive health information, we recommend CrystalDiskMark and CrystalDiskInfo. For cloning and imaging, the free Clonezilla is a good place to start.

PNY CS3150 — A Closer Look

Our review sample is the CS3150 1TB model with a black heatsink, with the optional RGB lighting. The drive is available in white or black, with or without RGB. The RGB version has a shorter heatsink underneath the fan shroud.

Speaking of fans, the two small fans that adorn the heatsink help cool the drive and are powered and controlled with a 4-pin connector on the non-RGB versions of the drive. If RGB is present, a 9-pin USB power connector is used instead, and the LEDs can be controlled through software.

The heatsink/shroud is listed as being 76 (L) x 23.5 (W) x 17.5mm (H), in case you have limited space to install a drive. Given the heatsink and fan requirements, the CS3150 effectively requires a desktop PC — putting one in a laptop or PS5 would require removal of the heatsink, at which point you'd be better off with a different drive, even if you want PCIe 5.0 speeds.

The heatsink can be removed if desired by removing the four screws and pulling the heatsink away from the drive. There are thermal pads on both sides of the drive — which makes sense, as it’s double-sided and the heatsink has a backplate (that might be easier to remove with some heat).

Not shown here, a thermistor or thermal sensor pokes through the top thermal pad to rest against the controller for temperature measurement. Thermistors change their resistance based on temperature, which allows the fans to vary in speed based on the controller temperature.

Looking at the back of the heatsink, we see that the drive is rated for 3.3V at 3.6A, which correlates to the maximum 11.55W the M.2 slot is designed to provide. Also listed is a physical security ID, or PSID, which enables a secure, data-destructive wiping of the drive in concert with the TCG Opal specification.

The drive’s DRAM is 2GB on the 1TB model and 4GB on the 2TB model. This is double the typical amount of memory, a fact we’ve covered in previous reviews. This is typical for drives based on Phison’s enterprise-inspired E26 SSD controller. DRAM is used for mapping and metadata storage, and in the case of enterprise applications or flash management, this extra memory could be put to good use.

MORE: Best SSDs

MORE: Best SSD for the Steam Deck

MORE: How We Test HDDs and SSDs

Comparison Products

The PNY CS3150’s primary competition includes other PCIe 5.0 drives of its caliber, as well as various PCIe 4.0 drives. This includes the Gigabyte Aorus Gen5 12000 and Nextorage NE5N. Since we’re looking at 1TB today, which is a harder sell for a high-end drive in comparison to 2TB, we’ve omitted the Teamgroup Cardea Z540 and Crucial T705, but do discuss where faster or larger drives might impact performance.

If you’re upgrading from a PCIe 4.0 SSD or are weighing 4.0 against 5.0, you’re probably using a drive based on Phison’s E18 SSD controller. In the lineup we have the DirectStorage-optimized Sabrent Rocket 4 Plus-G, the PS5-designed Seagate Game Drive, PNY’s own CS3140, and the popular Kingston KC3000, which is similar to the Fury Renegade. Also present is the OEM Transcend 250H that uses SMI’s comparable SM2264 controller, also found on the Adata Legend 960 and Legend 960 Max. Lastly, we’ve included two popular 1TB budget drives in the Teamgroup MP44L and Silicon Power UD90.

Trace Testing — 3DMark Storage Benchmark

Built for gamers, 3DMark’s Storage Benchmark focuses on real-world gaming performance. Each round in this benchmark stresses storage based on gaming activities including loading games, saving progress, installing game files, and recording gameplay video streams. Future gaming benchmarks will be DirectStorage-inclusive and we include details of that where possible.

The CS3150 performs quite well in 3DMark, beating all the other drives in the lineup. This includes the very similar Aorus Gen5 12000, or Aorus 12000 for short. Phison has improved the firmware on these drives over time, and drives can be optimized for certain benchmarks. It’s worth checking the results from other benchmarks to see if this is the case.

The CS3150, as an E26-based drive, does have DirectStorage-optimized firmware. So far this hasn’t meant much as few games use the technology. The most recent is Horizon Forbidden West, where load times can be somewhat improved over the PS5. The full API isn't utilized, however, so you probably won’t be getting any benefit over drives with newer flash like the Lexar NM790, which should have fast load times. If we’re talking full package, the Crucial T500 probably also gets you there at a lower cost.

Trace Testing — PCMark 10 Storage Benchmark

PCMark 10 is a trace-based benchmark that uses a wide-ranging set of real-world traces from popular applications and everyday tasks to measure the performance of storage devices.

The CS3150 also scores well in PCMark 10, making it the best drive in its class. This class includes E26 PCIe 5.0 drives like the Aorus 12000 and Teamgroup's Cardea Z540, which generally hit up to about 12,000 MB/s but don’t push the limit of the hardware. The optimized Max14um reference design, used on the ultra-fast Crucial T705, is faster still, but the T705 costs more — with or without a heatsink.

The point to take home here is that we’re looking at 1TB drives. It’s probably optimal to get these fast drives at 2TB or even 4TB so you get the most out of the platform, but if you’re looking for a fast 1TB option — and one that remains cool — then the CS3150’s results suggest it’s the drive to get.

Transfer Rates — DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with a custom, 50GB dataset. We write 31,227 files of various types, such as pictures, PDFs, and videos to the test drive, then make a copy of that data to a new folder, and follow up with a reading test of a newly-written 6.5GB zip file. This is a real world type workload that fits into the cache of most drives.

Things come back to earth once we move away from the UL benchmarks. The CS3150 is neck-and-neck with the Aorus 12000, with transfer speeds that outpace all PCIe 4.0 drives and the slower 5.0 drives as well. If you’re getting a drive for raw throughput, PCIe 5.0 can be worthwhile.

Synthetic Testing — ATTO / CrystalDiskMark

ATTO and CrystalDiskMark (CDM) are free and easy-to-use storage benchmarking tools that SSD vendors commonly use to assign performance specifications to their products. Both of these tools give us insight into how each device handles different file sizes and at different queue depths.

There are no surprises in ATTO, as the CS3150 follows the expected curves. Performance at 4KiB and up is good, which is generally what matters. It’s possible E26-based drives often have more DRAM to handle sub-page I/O, that is at block sizes less than 4KiB, which would indicate that the ATTO results at those block sizes suggest enterprise-like optimization. These values are lower than with other drives, which might be deceptive.

If we take a look at sequential performance in CDM, we see impressive numbers when compared to PCIe 4.0 drives. Earlier 5.0 drives, like the NE5N, are a little bit slower in comparison. If you’re wondering where a 1TB drive of this caliber would lose performance versus 2TB, it’s primarily with sequential reads and writes at high queue depth. That’s where the additional interleaving brings the most benefit. This is also true when you up the bus speed, as you would going from the CS3150 to the fully-fledged T705.

However, that bus speed can also come with better sequential writes at a low queue depth as you’re getting the most out of the flash’s potential. You don’t see the improvement with sequential reads at QD1 because reads cap out earlier than writes. While you would ideally have multiple threads going to get the most out of a PCIe 5.0 drive, basic transfers are at QD1, so if you’re looking for the best performance then it might be worth getting a faster drive. You would want a 2TB drive to get better maximum throughput. The 1TB CS3150 is not a terrible solution if you’re pulling sequential reads, though, which applies to traditional tasks like game loads.

Sustained Write Performance and Cache Recovery

Official write specifications are only part of the performance picture. Most SSDs implement a write cache, which is a fast area of (usually) pseudo-SLC programmed flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the cache and into the "native" TLC or QLC flash.

We use Iometer to hammer the SSD with sequential writes for more than 15 minutes to measure both the size of the write cache and performance after the cache is saturated. We also monitor cache recovery via multiple idle rounds. This process shows the performance of the drive in various states as well as the steady state write performance.

The 1TB CS3150 writes in pSLC mode at just over 9.3 GB/s for about 12 seconds. This reveals a relatively small cache of about 112GB, which is about one-third the possible maximum for a drive of this size. It’s still sufficient for most tasks, and the use of a smaller cache can allow for better consistency.

The direct-to-TLC mode writes at over 1.8 GB/s as indicated by the steady state. This is exactly as expected and is a good, consistent result. It’s not any faster than what was achievable with fast PCIe 4.0 drives, but that’s an expected compromise for the first round of 5.0 SSDs. If you're aiming for sustained write performance specifically, you can probably settle for a good 4.0 SSD.

Power Consumption and Temperature

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is an important aspect to consider, especially if you're looking for a laptop upgrade as even the best ultrabooks can have mediocre stock storage. Desktops are more performance focused with less support for power-saving features, so we show the worst-case idle power use here.

Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption but performance-per-watt, or efficiency, is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state more quickly, ultimately saving energy.

For temperature recording we currently poll the drive’s primary composite sensor during testing with a 21–22°C ambient. Our testing is rigorous enough to heat the drive to a realistic ceiling temperature.

The CS3150 is not an efficient drive, matching the Aorus 12000 almost exactly. Most PCIe 4.0 drives are more efficient, and in some cases much more efficient. Still, as the effective bus speed increases on E26-based drives, efficiency improves. This is seen here when comparing the CS3150 to the NE5N. The 1TB T705, if present, would be even more efficient.

If you’re looking for power efficiency, though, you are best off going with PCIe 4.0 or waiting for the next generation of 5.0 SSD controllers. The CS3150 is not intended for use in a laptop, so this isn’t a huge deal, though idle power draw in most desktops is still high.

The CS3150’s cooler is excellent, keeping the drive below 50°C during our write saturation testing. This drive won’t be overheating any time soon. That said, you might want to adjust the fan curve to produce a quieter state. A ceiling of 75°C or so is reasonable for any SSD, and the heatsink alone on the CS3150 is quite capable.

Test Bench and Testing Notes

We use an Alder Lake platform with most background applications such as indexing, Windows updates, and anti-virus disabled in the OS to reduce run-to-run variability. Each SSD is prefilled to 50% capacity and tested as a secondary device. Unless noted, we use active cooling for all SSDs.

PNY CS3150 Bottom Line

The PNY CS3150 is a very fast drive, and that alone earns it a decent score. Performance is good to great in all benchmarks, and sustained performance is also quite good for a 1TB drive. The drive should be even faster at 2TB, and that’s usually what you'd want for a PCIe 5.0 drive. On the other hand, if you only need 1TB of fast storage, then this drive will get the job done. You won’t be losing much versus 2TB in many workloads, and you get the benefits of DirectStorage-optimized firmware, TCG Opal 2.0 support, and throughput that outmatches any PCIe 4.0 drive.

DirectStorage optimization is a hard sell, without much current use. Even if Phison’s firmware was a selling point, drives like the Crucial T500 would be a less expensive option. Hardware encryption is nice, but it’s more popular on laptops and this drive is not designed for such devices. More throughput is always welcome, but it can be difficult to use in the real world and there are faster drives like the Crucial T705. This leaves the CS3150 feeling a bit like a compromise aimed particularly at those who want the drive’s dual-fan cooler and, optionally, the RGB lighting.

The drive does run incredibly cool for its level of performance and, thankfully, the fan speed is customizable. Even with the fans off, the heatsink helps keep the drive from throttling. For the RGB fanatics, especially those with other PNY products, the ability to control and sync the LEDs is a nice feature. If you don’t like dealing with extra wires, though, you might be better off a standard PCIe 5.0 drive that has a passive heatsink or allows you to use your own, like the Crucial T700 or T705. Or, at this specific performance level, the Teamgroup Cardea Z540.

Many users would be better off with a 4.0 SSD, like the WD Black SN850X or Samsung 990 Pro, for general performance. For sustained performance, drives like the Seagate FireCuda 530 can be an excellent choice. This makes the CS3150 feel a bit underwhelming unless you’re set on the cooler and/or RGB. These features aside, this drive could be a good alternative if it’s priced right. Unfortunately, at the time of review its pricing is not very compelling, and availability has been spotty with the 2TB being currently unavailable. There’s also no 4TB option, although that has been hard to find in any PCIe 5.0 drive. The CS3150 ends up being fast, but with a very niche clientele in the current market.

MORE: Best SSDs

MORE: Best SSD for the Steam Deck

MORE: How We Test HDDs and SSDs

The Nvidia RTX 4070 Ti Super officially launched on January 24, with base model cards starting at $799. Nvidia sent us a reference clocked Asus TUF Gaming for our initial launch review, and PNY sent its Verto OC — or Verto Epic-X RGB OC if you want the full name, though we'll use the shorter description throughout this review. This will show what a minor factory overclock can provide, and as you can guess there's plenty of RGB, so it's priced $50 higher at $849. There's no Founders Edition for the 4070 Ti Super, leaving it to the add-in board (AIB) partners to fight it out for supremacy and a place among the best graphics cards.

There are plenty of other reference clocked and factory overclocked cards — check our list of all the RTX 4070 Ti Super cards — and we can expect the usual RGB blinged-out models and larger coolers, as well as perhaps some smaller offerings. PNY usually takes a more conservative approach, or at least it has in the past, but the latest Verto OC designs are starting to include quite a bit of RGB.

Here's a quick look at the specs for the various GPUs, truncated as this is a third-party follow-up review. Check the full launch review and the GPU benchmarks hierarchy for additional details.

The only real difference in terms of specs between the PNY Verto OC and the reference 4070 Ti Super comes from the small 45 MHz bump in boost clock. On paper that's a 1.7% increase, though there are other factors as well. TGP is the same 285W, as far as we're aware, but the cooler, fans, VRMs, and other elements can all have a knock-on effect.

There shouldn't be any major changes in overall performance, and we won't spend a ton of time analyzing the results as they mostly speak for themselves, but let's take a closer look at the PNY card itself before hitting the benchmarks.

PNY RTX 4070 Ti Super Verto OC

First looks don't suggest anything unusual for the PNY Verto OC. It's a larger card, measuring 333x134x64 mm (our measurements), and weighing 1547g. That's close in dimensions to the Asus TUF Gaming we looked at for the launch review, except the PNY card is about 3cm longer and weighs 224g more. Both are slightly wider than a 3-slot width and will basically occupy four slots.

PNY also uses different fans, and this goes along with the potentially improved cooling performance. There are three 98mm fans, with the integrated rims that we like to see on modern graphics cards. That extra centimeter in diameter should mean more airflow and lower temperatures, and our test results on page five will confirm that.

The difference in weight could be due to a better cooler in theory, but more likely is that the larger fans plus the full 3-slot IO bracket account for most of that weight. There's also the materials and construction of the cooling shroud and RGB lighting that could account for at least a bit of weight.

PNY currently lists two different 4070 Ti Super variants: the Verto OC we're looking at includes full RGB lighting, while the regular Verto only has a 15 MHz factory OC, a slightly thinner cooler, and rim-less fans. It probably works fine, but we definitely appreciate the extras on the Verto OC.

RGB lighting on the card consists of two main areas. There's a large "X" on the front of the card that lights up, and then there's an XLR8 logo on the top of the card. Both can be controlled by PNY's Velocity software, or left to their default color cycling behavior.

The only extra in the packaging consists of a support stand — not one of the simple pre-assembled models. The stand does work, but it's not quite as convenient as other solutions, with discrete heights that can be set via the various screw holes, and then a slider on the small support bracket for finer adjustments.

PNY sticks with the usual triple DP1.4a ports and a single HDMI 2.1 output for the display options. All outputs can support up to at least 4K and 240 Hz with HDR, using Display Stream Compression.

Here's a quick comparison photo of the two RTX 4070 Ti Super cards we've tested. The PNY card on the right looks longer and thicker, but the two are actually close to the same thickness — it's just the extra length making it look thicker. We do like the inclusion of a full 3-slot IO bracket for added support of the card, since it blocks the use of all three expansion slots, though for cases with vertical mounting options sometimes it's better to have a 2-slot bracket. (Most cases that have a vertical mounting option don't have a 3-slot cutout.)

PNY RTX 4070 Ti Super Verto OC Test Setup

Our current graphics card testbed has been in use for over a year now, and so far we haven't seen any pressing need to upgrade. The Core i9-13900K is still holding its own, and while the i9-14900K or Ryzen 9 7950X3D can improve performance slightly, at higher resolutions and settings we're still almost entirely GPU limited — though perhaps not when the future RTX 50-series and RX 8000-series arrive.

We're using the same preview drivers from Nvidia, version 551.15, for our testing (which was done in advance of the 4070 Ti Super launch). In our 4070 Super testing, we noticed some changes in several of the benchmarks and retested the Nvidia cards in the affected games. We've also retested AMD's GPUs using the 23.12.1 drivers. The most impacted games are Borderlands 3, Far Cry 6, Forza Horizon 5, Microsoft Flight Simulator, and Spider-Man: Miles Morales. We've also fully retested the RTX 4070 and RTX 4070 Ti.

Since this is a third-party post-launch review, we'll omit the extra testing that we conduct on launch cards, focusing just on gaming performance, cooling, and other aspects of the cards. We've included eight other GPUs from AMD and Nvidia as reference points for this review, and we'll highlight the Asus and PNY cards in the charts.

Our current test suite consists of 15 games. Of these, nine support DirectX Raytracing (DXR), but we only enable the DXR features in six games. The remaining nine games are tested in pure rasterization mode. We don't enable upscaling on any of the cards for our primary tests, and we run 1080p (medium and ultra), 1440p ultra, and 4K ultra benchmarks — ultra being the highest supported preset if there is one, and in some cases maxing out all the other settings for good measure (except for MSAA or super sampling).

Our PC is hooked up to a Samsung Odyssey Neo G8 32, one of the best gaming monitors around, allowing us to fully experience the higher frame rates that might be available. G-Sync and FreeSync were enabled, as appropriate. As we're not testing with esports games, most of our performance results are nowhere near the 240 Hz limit, or even the 144 Hz limit of our secondary test PC.

We've installed Windows 11 22H2 and used InControl to lock our test PC to that major release for the foreseeable future (though critical security updates still get installed monthly — and one of those probably caused the drop in performance that necessitated retesting a third of the games in our suite).

Our test PC includes Nvidia's PCAT v2 (Power Capture and Analysis Tool) hardware, which means we can grab real power use, GPU clocks, and more during all of our gaming benchmarks. We'll cover those results on our page on power use.

Finally, because GPUs aren't purely for gaming these days, we've run some professional content creation application tests, and we also ran some Stable Diffusion benchmarks to see how AI workloads scale on the various GPUs.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

The RTX 4070 Ti Super makes the most sense to us as a 1440p gaming solution — which is a close stand-in for 4K with DLSS Quality mode upscaling. So that's where we'll start our benchmarks.

We have 15 games and 18 charts, with the "high level overview" showing the geometric mean across all 15 games. Then we break things into rasterization and ray tracing test suites on the next two charts, followed by the 15 individual gaming results.

As this isn't the primary launch review, we'll just put all 18 charts into a single gallery and then discuss the results.

The main question we're looking to answer here is: What does the PNY card compared to a reference clocked card? The answer, as you can probably guess, is that it's not much. PNY's Verto OC consistently comes in a few percent ahead of the Asus card, but there's nothing that would be readily noticeable in normal used.

Overall, at 1440p the PNY card was 1.8% faster than the Asus card. In just the rasterization suite, that difference increased to 2.0%, and in ray tracing it dropped to 1.5%, but those are all practically within margin of error. Still, the PNY card is technically faster, and other aspects of the card might catch your eye.

Even if we look at the individual game results, there's not much to see. Across the full suite, the PNY card leads the Asus card by anywhere from 0.3% to 3.1%, with no clear indication of why certain games benefit more. We do run each test multiple times, discarding the first run and taking the best result of the remaining runs (after checking that the runs are reasonably consistent), but even then there's a bit of wiggle room.

As for the rest of the GPUs, our initial review covered the details, and there's nothing really new to report. A 2% improvement isn't enough to catch the RTX 4080, and the PNY card basically splits the difference between the vanilla 4070 Ti and the 4080. It's 9.2% faster than the 4070 Ti and 9.6% slower than the RTX 4080. The small improvement did allow the PNY card to squeak past the AMD RX 7900 XTX in the overall chart, though of course the individual games tell the full story: Faster DXR from Nvidia, faster rasterization from AMD.

We should note that, while we try to use reference cards whenever possible, both the 4070 Ti and 4070 Ti Super only shipped from third-party AIB partners. We used reference clocked Asus cards for both the "standard" models.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

4K — particularly with upscaling (from 1440p) — also represents a reasonable target for any graphics card that costs $800 or more. What does the PNY Verto OC provide that the Asus TUF Gaming doesn't? Mostly RGB lighting, but let's look at the 4K performance.

The results pretty much match what we saw at 1440p, with the PNY card leading the Asus card by 2.2% overall. That doesn't change any of the overall rankings, at least not in any meaningful way. The 7900 XTX just edges past the PNY card, with a decently large 16.9% lead in the rasterization suite overcoming the 17.1% deficit in the DXR suite — but there are more rasterization than DXR games, so overall AMD still comes out on top.

Getting back to the two 4070 Ti Super cards, PNY increases its lead just slightly at 4K, ranging from 0.8% to 3.9% faster in the 15 individual games. Curiously, the biggest lead was in Flight Simulator, which is normally more CPU limited. Looking into the details, it seems both cards were at the TGP limit, but the PNY card was running 5C cooler and that allowed it to average 100 MHz higher on its clocks — 4% higher than the Asus card's clocks. We'll get into this more on page five, but cooling certainly looks to be a strong point on the PNY Verto OC.

One thing that's clear from our 4K test suite is that ray tracing at native 4K proves to be too much for the 4070 Ti Super. Games that aren't quite as demanding, like Spider-Man, can get close to 60 fps, but Bright Memory Infinite (Benchmark) and Cyberpunk 2077 would need upscaling and possibly frame generation to crack 60 fps. The standalone benchmark for BMI doesn't have DLSS 3 support, but the main game does, if you're wondering — and the main game is far less taxing than the impressive looking tech demo.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

Wrapping things up, we have both 1080p medium and 1080p ultra testing results. Medium isn't necessarily an important consideration, and 1080p in general isn't really where we think most gamers spending this much for a GPU are going to play. Still, we ran the benchmarks so here are the charts.

Because 1080p gaming puts the bottleneck more on the CPU than the GPU, particularly with high-end cards like the 4070 Ti Super, the relative margins all shrink a bit. The PNY card ends up being 1.5% faster than the Asus card at 1080p ultra, and only 1.2% faster at 1080p medium — again, not something you'd notice or feel in gaming, even if it's technically visible in our benchmarks.

Several games are completely CPU limited now, like Far Cry 6, Flight Simulator, Horizon Zero Dawn, and even Spider-Man: Miles Morales and Watch Dogs Legion. That gives us a range of -0.4% slower to 3.0% faster at 1080p, with the usual ~0.5% margin of error. If we only look at 1080p medium, several more games become effectively CPU limited, which you can see by the overall rasterization chart where the top five GPUs are only separated by 4%.

Running all of these benchmarks on these GPUs might seem a bit pointless. A quick look at the specs tells most of what we find in the charts. But there's a method to the madness, because we also want the data on power use, clock speeds, and temperatures that we cover on the next page.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

Our graphics card test PC uses an Nvidia PCAT v2 device, and we've switched from the Powenetics hardware and software we've previously used to Nvidia PCAT, as it gives us far more data without the need to run separate tests. PCAT with FrameView allows us to capture power, temperature, and GPU clocks from our full gaming suite — along with the usual frame times. The charts below are the geometric mean across all 15 games, though we also have full tables showing the individual results further down the page.

The raw performance data in the games can sometimes be interesting, particularly when we're looking at a brand-new GPU. For these third-party card reviews, however, outside of aesthetics we find the noise and temperature results to be the most useful information we can glean from all the testing.

We have separate charts for 1080p medium/ultra, 1440p ultra, and 4K ultra in the galleries below. While the Asus and PNY 4070 Ti Super cards were very close in frames per second, we start to see a bit more differentiation in the other aspects of the cards.

PNY RTX 4070 Ti Super Verto OC Power Use

PNY's slight factory overclock at least partly explains the performance difference, but what's also interesting is that it's able to provide that slightly higher performance while using less power. Temperatures can and do impact power use a bit, but overall the key takeaway is that PNY's Verto OC averaged 9–10 watts less across all four test resolutions/settings.

That's a 3.2% to 4.5% reduction in power use, with about a 2% increase in performance, meaning overall efficiency (FPS/W) improves by about 6%. That's not a massive increase, but it's still measurable and interesting, especially when you consider there's quite a bit more RGB lighting on the PNY card.

Better efficiency and lower power use should mean less noise as well, but let's check the other aspects first.

PNY RTX 4070 Ti Super Verto OC Clock Speeds

The official boost clock on the RTX 4070 Ti Super is 2610 MHz, and the PNY card has a 2655 MHz boost clock. That's a 1.7% increase, which basically matches the performance changes we measured, but there's a bit more to it than stated boost clocks.

Looking at the real-world clocks, depending on the resolution/setting, the PNY card averaged 2828–2893 MHz while the Asus card averaged 2717–2776 MHz. That would work out to a 4% advantage in theory, but CPU, memory, and system bottlenecks are still a factor.

PNY RTX 4070 Ti Super Verto OC Temperatures

One area where the PNY Verto OC excels that we've already hinted at is its temperatures. Large coolers and large fans produce lower temperatures? Shocking! But even if it seems like a relatively simple approach, average temps on the PNY ranged from just 48C at 1080p medium to 54C at 4K ultra. More importantly, that's about 9C lower than the Asus card that uses the same GPU and runs at slightly lower clocks.

These really are exceptional cooling results, considering the 285W power limit. Across our entire test suite, we never saw the GPU average more than 57C in any of the games. But it's not just about the temperatures, as we also need to check noise levels.

PNY RTX 4070 Ti Super Verto OC Noise Levels

The PNY Verto OC ends up just slightly quieter than the Asus card, which is good to see. Spinning fans at higher RPMs in order to get lower temperatures isn't a great tradeoff. It's not a silent card, but 40–41 dB(A) at a close range isn't very loud either. This was with a fan speed of 30%, which brings up something more to discuss.

One slight oddity, if you want to call it that, is that the PNY fans go from off to 30% once the GPU gets above about 40C. We've seen other cards where the minimum fan speed is 20%, which means they can tune for lower noise if desired while still keeping thermals in check. Even with overclocking software, we were unable to set the PNY fans below 30%. So, the overall noise levels while gaming are good, but in lighter workloads, the card will be just as loud as while gaming.

We also tested with a static fan speed of 75%, which resulted in the PNY card generating 56.7 dB(A) of noise. That's very noticeable, but it's also very unlikely you'd ever get the card hot enough to need such a high fan speed. Even in warmer climates, we suspect the fan will still be able to run at 30–40 percent while keeping the GPU below 70C — unless you're gaming in a room that's over 40C, I suppose, but that's not something I would ever recommend doing.

Our noise test consists of running Metro Exodus, because it's one of the more power hungry games. We load a save, with graphics set to appropriately strenuous levels (4K ultra in this case), and then let the game sit for at least 15 minutes before checking noise levels. We place the SPL (sound pressure level) meter 10cm from the card, with the mic aimed at the center of the middle fan (or the center of the rear fan if there are only two fans). This helps minimize the impact of other noise sources, like the fans on the CPU cooler. The noise floor of our test environment and equipment is around 31–32 dB(A).

Full Gaming Test Results

For those who like to look at the finer details, we've provided the above tables showing all of our gaming benchmarks, which include FPS/$ and FPS/W metrics. We use the best prices we could find at the time of writing: $799.99 as the base MSRP for the Asus card and $849.99 for the PNY Verto OC — though note that both are currently out of stock at Newegg, Amazon, and the other places we checked.

As noted already, the PNY card ends up being more efficient than the Asus card, though it's not quite able to match the 4070, 4070 Super, or 4080 Founders Edition cards that we've tested.

FPS/$ is a different story. Paying $50 extra represents a 6.25% price increase relative to the base model $800 cards. Since you're only getting 2% more performance, give or take, that means it's a "worse" overall value. But of course it's not just about the performance, and many a gamer has willingly paid extra to get a card aesthetic that they like.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

If the only thing you care about when it comes to choosing a graphics card is the raw performance, the advice is simple: Buy the cheapest card you can find that has the GPU you want. Nvidia's RTX 4070 Ti Super is a good choice overall, and whether you get a card from PNY, Asus, MSI, Gigabyte, Zotac, or some other vendor, it should deliver very nearly the same level of performance as the cards we've tested — give or take a few percent. But there's more to picking a graphics card than pure performance for a lot of gamers.

Objectively, cooling performance, noise levels, and efficiency are measurable characteristics where the PNY Verto OC excels. We've only tested two different 4070 Ti Super cards at present, and some of the other options might match or even beat the PNY, but all indications are that it's a great option if you value lower GPU temperatures, reduced noise levels, and a slightly lower power draw.

Subjectively, appearance is also a big factor for a lot of people. We've provided images and will leave opinions on how the card looks mostly up to potential buyers, but it's a decent looking traditional triple-fan design in our book. It doesn't try anything radically different from other graphics cards, so there are no swooping curves or wild color schemes, and that should be fine for a lot of people.

PNY has been around a long time in the graphics card market, though much of that time was spent primarily as a provider of Nvidia reference designs. It's shifted away from that in recent years, with more of an emphasis on creating competitive enthusiast designs. The Epic-X RGB Verto OC isn't at the same hardware tier as Asus Strix or MSI Suprim offerings, but it's priced more competitively and provides everything most people could want.

The bigger issue it faces is that first people need to be in the market for a high-end graphics card, and not already have taken the plunge with an RTX 4080, RTX 4070 Ti, RX 7900 XTX, or RX 7900 XT. It's also $50 extra for things like RGB lighting and extra cooling capabilities that aren't strictly necessary.

Fundamentally, the Nvidia RTX 4070 Ti Super represents a move in the right direction as far as price and performance go, compared to the existing RTX 4080 and RTX 4070 Ti, but that doesn't mean the GPU actually gets to where most people would like to see it: below $700. It probably never will before it gets superseded by a next generation part.

The PNY RTX 4070 Ti Super Epic-X RGB certainly warrants consideration if you're in the market for a high-end GPU. It can compete with AMD's best, with the usual caveats that it's comparatively slower in rasterization and much faster in ray tracing. It's also one of the coolest running high-end GPUs of recent memory. It doesn't break new ground, but it didn't need to.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

While there's definitely an argument to be made about RGB overload, there's no doubt that colored lights can make a component stand out, whether on the shelf or in your system. So it's no surprise that PNY is trying to appeal to the masses by giving you some aesthetic options with its latest XLR8 CS3150 PCIe 5.0 NVMe SSD.

These sticks are cooled by two fans on both ends of the heatsink, and come in either black, white or an Epic-X RGB backlight version. The latter in particular looks like a tiny graphics card. 

The drive is 39.8mm (1.56 inches) tall, which makes it a whole lot friendlier to other components in your case than the Nextorage NE5N. The drive is actually fairly compact, given it houses a backplate, two slices of the thermal pad with the drive in between, the heatspreader and, with its RGB variant, a shroud.

The 1TB version is rated to 11,500 MB/s and 8,500 MB/s sequential read and write speeds, with the 2TB rated up to 12,000 MB/s and 11,000 MB/s. Its fans draw power via a 4-pin power connector. Hopefully that's referring to a 4-pin fan headers plug and not something proprietary or as ancient as Molex. The specs sheet and images don't mention specifics there. The Inland TD510 which uses the same controller includes a Molex/ SATA Power connector to power up its fan. If it does use headers, it would be nice to see if variable fan speeds can be controlled or if it makes any difference at lower RPM, hence lower noise. 

Here are the drive's full specs:

With active cooling increasingly included on these high-end SSDs, manufacturers should also include fan specs such as bearing type, CFM and the max dB rating. This is standard practice with fans, CPU and GPU cooler makers.

The Tall Tale of Tall Heatspreaders, Waterblocks and Fans

So far, all PCIe 5.0 SSDs with Phison chips are either provided with passive cooling such as the Aorus Gen 5 and the Crucial T700, or with a heatsink pre-installed and a tiny fan like Micro-Center's in-house brand Inland TD510. As we saw in our review of the Inland drive, the fan didn't actually improve cooling. Naturally, it depends on how the cooling is designed to complement the chips and their layout, as well as how much air is already circulating in your case and around your M.2 slot.

But one of the appealing aspects of an M.2 drive has been its thin form factor, which can hidden away under motherboard shrouds,  like a six-M.2-slot ASRock Z790 Nova WiFi. But drives like these with tall heatsinks and fans have to live loud and proud, sticking up off your motherboard.

This issue, though, is is largely down to the Phison E26 controllers at the heart of these drives. As was mentioned heavily in our Corsair MP700 review, there is a lot of work to be done on the efficiency front with PCIe 5.0 SSDs. In the meantime, Corsair sells a liquid cooling block for this 2280 form factor NVMe drive, which reminds me of the days of squeezing performance from Intel's Prescott Pentium 4 processors, which were inefficient  for their time.

We should see new PCIe 5.0-based controllers in the following months and hopefully quicker, cooler and quieter SSDs hitting store shelves. Just a day ago, a Silicon Motion SM2508 controller showed promising efficiency and performance. There's also the TeamGroup T5 PCIe 5.0 NVMe that will use an InnoGrit G5666 controller.

We'll likely learn more about how efficient these drives are early next year at CES 2024 or not long after. So if you're not keen on a PCIe 5.0 drive due to fans and bulky coolers, you might not have long to wait for better options. Just don't expect them to be cheap when all signs are pointing to SSD price increases next year.

August 25, 2023 Update: We've added new testing for the 2TB PNY XLR8 CS3140 SSD on page 2.

Original Review published June 22, 2022:

PNY has been around for a long time, known in particular for its flash-based products, but also for its memory and GPUs. The company's new XLR8 CS3140 SSD tops out at 7,500/6,850 MBps of sequential read/write throughput for the 2TB and larger models, and it's also flexible in that it has both a version with a heatsink and one without, and you can also use a separate PlayStation 5 heatsink for the latter. 

Last year, PNY reduced the endurance rating on its CS3030 SSD by almost 80% due to the impact of Chia cryptomining, which was known for killing SSDs with heavy write workloads. But PNY has now returned to high endurance ratings with the XLR8 CS3140.

We tested the new 1TB XLR8 CS3140 model, which is a good starting capacity, but the 2TB model offers better peak performance because it is packed with more flash dies. Naturally, the 4TB is even more compelling if you're after more raw capacity. However, a basic primary drive is quite fine at 1TB, especially when coupled with strong hardware and a good SLC caching scheme like we see with the previous-gen CS3140. Let's see how the PNY CS3140 fares in our test suite. 

Specifications

The PNY CS3140 is available in the 1TB, 2TB, and 4TB capacities. These all have a five-year warranty backed up by 700TBW, 1400TBW, and 3000TBW of write endurance, respectively. Performance tops out at 7,500/6,850 MBps of sequential read/write throughput for the 2TB and larger models.

PNY doesn't officially list the IOPS measurements, but we can guess that this drive would be similar to other SSDs that have the same hardware combination. We had to dig up the endurance numbers on PNY's Taiwanese site, and the company seems to operate on the principle of “the less the information, the better.” That does make some sense in an era where SSD manufacturers often change the SSD's hardware after launch, but you'll need to do your homework to find concrete specs for these drives.  

The drive supports 256-bit AES encryption. PNY sells this drive with and without a heatsink, and we are testing the version without. This makes it a good choice for the PlayStation 5, possibly as-is, but PNY sells a separate heatsink for SSDs that is designed especially for the PS5. 

Software and Accessories

The CS3140 doesn't come with anything special. PNY does offer a download for its PCIe SSD toolbox software for firmware updating on their site. This software also gives some basic information about the drive, such as SMART readings, and allows for secure erasing.

A Closer Look

The drive is quite plain, sporting a rear label with basic details about the drive. Under the top label, we see the controller, DRAM, and four NAND packages in a standard configuration.

The flash is labeled IA7BG94AYA, which is 176-layer Micron TLC, a.k.a. B47R. This flash has proven excellent when coupled with the Phison E18 controller. As there are four packages, each should have four 64GB flash dies in a QDP configuration.

Based on data pulled from the drive, this flash is standard B47R running at 1200 MT/s. We’ve seen faster, but this is sufficient with two dies per channel. The endurance and quality of the flash look good, so there are no surprises here.

The Phison E18 controller needs no introduction - it’s been used on a wide variety of drives to good effect. It tends to be optimized for peak performance, specifically in sequential work, and it’s particularly good when paired with newer flash.

The DRAM is SK hynix’s H5AN8G6NCJR-VKC, which is 1GB of DDR4 in the 512M x 16b configuration. This would be C-die DRAM with the normal DRAM-to-NAND ratio, more than enough for any user. We can also spot the Phison power management package (PMIC) near the controller.

MORE: Best SSDs

MORE: Best SSD for the Steam Deck

MORE: Best Hard Drives

MORE: Best External SSDs

Comparison Products

There are some new drives out on the market that weren’t here when we reviewed the 1TB CS3140, including PCIe 5.0 SSDs like the Crucial T700 and Corsair MP700. We also have reviewed some great 4.0 SSDs like the Samsung 990 Pro, the Solidigm P44 Pro/SK hynix Platinum P41, and the WD Black SN850X. There’s also the Sabrent Rocket 4 Plus-G and Adata Legend 960 Max. Lastly, we have the budget-oriented Solidigm P41 Plus. Keep in mind that we got the PS5 heatsink with the 2TB CS3140 which makes it great for the PS5 even if the drive is dated in standard benchmarks.

A Closer Look

Trace Testing - 3DMark Storage Benchmark

Built for gamers, 3DMark’s Storage Benchmark focuses on real-world gaming performance. Each round in this benchmark stresses storage based on gaming activities including loading games, saving progress, installing game files, and recording gameplay video streams.

The CS3140 is below average in 3DMark. It scored slightly worse than the 1TB version.

Trace Testing – PCMark 10 Storage Benchmark

PCMark 10 is a trace-based benchmark that uses a wide-ranging set of real-world traces from popular applications and everyday tasks to measure the performance of storage devices.

The CS3140 is near the bottom in PCMark 10, too, faring worse than the 1TB version.

Transfer Rates – DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with a custom, 50GB dataset. We copy 31,227 files of various types, such as pictures, PDFs, and videos to a new folder and then follow-up with a reading test of a newly-written 6.5GB zip file.

Things aren’t quite as bad with DiskBench. The CS3140 is close enough to the other high-end PCIe 4.0 SSDs. The 5.0 SSDs - the T700 and MP700 - pull away due to their faster interface, while the slower P41 Plus is at the bottom.

Synthetic Testing - ATTO / CrystalDiskMark

ATTO and CrystalDiskMark (CDM) are free and easy-to-use storage benchmarking tools that SSD vendors commonly use to assign performance specifications to their products. Both of these tools give us insight into how each device handles different file sizes.

The CS3140’s sequential performance in ATTO and CDM is adequate but not on the level of the better PCIe 4.0 SSDs. This is fine as sequential performance at this level isn’t super important. Random write latency is good — surprisingly good even, but random read latency is fairly bad and significantly worse than the 1TB version of this drive.

The reason for that is obvious; PNY switched to 112-Layer Kioxia TLC (BiCS5) which does not perform as well with this controller as the original 176-Layer Micron TLC (B47R). However, if you’re buying a drive for capacity and especially for the PS5, this is not a big deal.

Sustained Write Performance and Cache Recovery

Official write specifications are only part of the performance picture. Most SSDs implement a write cache, which is a fast area of (usually) pseudo-SLC programmed flash that absorbs incoming data. Sustained write speeds can suffer tremendously once the workload spills outside of the cache and into the "native" TLC or QLC flash.

We use Iometer to hammer the SSD with sequential writes for 15 minutes to measure both the size of the write cache and performance after the cache is saturated. We also monitor cache recovery via multiple idle rounds.

The 2TB CS3140 writes in pSLC mode for over 30 seconds at up to 6.2 GB/s. This makes for a cache over 200GB, which is fairly large. Luckily, the drive does not easily hit a folding state, and instead drops to around 1.6 GB/s in TLC mode. This is not a bad result in any way. The 1TB version with Micron TLC is faster in TLC mode and it would probably be even faster with more dies at 2TB. Still, the BiCS5 2TB CS3140 is fairly consistent. Its recovery response has it lingering in TLC mode, which is luckily not too bad.

Power Consumption and Temperature

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is an important aspect to consider, especially if you're looking for a laptop upgrade as even the best ultrabooks can have mediocre storage.

Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state more quickly, ultimately saving energy.

For temperature recording we currently poll the drive’s primary composite sensor during testing with a 24C ambient.

The 2TB CS3140 with BiCS5 flash is not very efficient, especially compared to similar drives using Micron’s TLC instead. This is to be expected as this flash is less efficient in general and does not seem to pair quite as well with Phison’s E18 controller. This result isn’t crtical if this drive is used in a high-end desktop or PS5, but it does demonstrate the downsides of swapping the flash. It’s worth pointing out that other manufacturers can and have changed hardware over the course of an SSD’s lifespan.

We only reached a brief maximum of 67C during testing, which is far away from significant throttling. This drive would be quite happy in a PS5 with the optional heatsink. We would recommend it for that or perhaps some extra fast storage if the drive is priced right. Right now, at the time of the review update, the 2TB SN850X is probably the better option.

Test Bench and Testing Notes

We use an Alder Lake platform with most background applications such as indexing, Windows updates, and anti-virus disabled in the OS to reduce run-to-run variability. Each SSD is prefilled to 50% capacity and tested as a secondary device. Unless noted, we use active cooling for all SSDs.

MORE: Best SSDs

MORE: Best SSD for the Steam Deck

MORE: Best Hard Drives

MORE: Best External SSDs

PNY XLR8 CS3140 1TB SSD

RATING: ★★★★ (4 Stars)

PROS

+ Consistent performance
+ Strong peak and all-around performance
+ Power-efficient
+ Good PlayStation 5 option from PNY

CONS

- Pricing

OUR VERDICT
The 1TB PNY CS3140 is a solid drive that can be used in a PC or in the PlayStation 5, but the latter is best with PNY’s separate PS5 heatsink. Performance is strong, but there’s not much special about this drive versus the competition.

Comparison Products (1TB)

We compared the PNY CS3140 to other PCIe 4.0 drives, including the more budget-oriented WD Black SN770. Contenders using proprietary controllers include the WD Black SN850, the Samsung 980 Pro, and the Crucial P5 Plus. The other three drives use the same Phison E18 controller, but the Gigabyte Aorus Gen4 7000s and Corsair MP600 Pro use the older 96-layer Micron TLC while the last one, the Kingston Fury Renegade, is equipped with the same 176-layer flash as the CS3140.

Trace Testing - 3DMark Storage Benchmark

Built for gamers, 3DMark’s Storage Benchmark focuses on real-world gaming performance. Each round in this benchmark stresses storage based on gaming activities including loading games, saving progress, installing game files, and recording gameplay video streams.

3DMark shows off WD’s excellent optimization, but thanks to their upgraded flash, we also had respectable results for the CS3140 and Fury Renegade.

Trace Testing – PCMark 10 Storage Benchmark

PCMark 10 is a trace-based benchmark that uses a wide-ranging set of real-world traces from popular applications and everyday tasks to measure the performance of storage devices.

Again we see strong optimization from WD and Crucial, with the newer flash helping out the CS3140 and Fury Renegade. PCMark is often considered a “real world” benchmark, and we can look particularly at the latency results to see that the older 96-layer flash on the 7000s and MP600 Pro is appreciably slower.

Transfer Rates – DiskBench

We use the DiskBench storage benchmarking tool to test file transfer performance with a custom, 50GB dataset. We copy 31,227 files of various types, such as pictures, PDFs, and videos to a new folder and then follow-up with a reading test of a newly-written 6.5GB zip file.

This bandwidth-heavy test favors the high-end PCIe 4.0 drives, particularly those based on the Phison E18 controller. The newer 176-layer flash also brings improvements, particularly as we are testing at 1TB rather than 2TB of capacity. The Phison E18 can peak with 32 dies of flash, and considering these use 64GB dies, that means 2TB. The 176-layer flash is also faster per die than the 96-layer flash for copying tasks.

Synthetic Testing - ATTO / CrystalDiskMark

ATTO and CrystalDiskMark (CDM) are free and easy-to-use storage benchmarking tools that SSD vendors commonly use to assign performance specifications to their products. Both of these tools give us insight into how each device handles different file sizes.

The queue depth 1 (QD1) sequential read and write results from ATTO paint a simple picture: drives using the Phison E18 come out on top, with a significant edge given to those with the newer 176-layer flash. To some extent, there is a “wall” due to the limits of the x4 PCIe 4.0 interface, as seen with queue depth 8 sequential reads in CrystalDiskMark. The write result shows a bit more delineation between the tested drives. At this point, we see a theme, with the CS3140 underperforming the Fury Renegade, possibly due to inferior firmware optimization.

Random QD1 performance is good to excellent on the CS3140, where we again see the newer flash pulling ahead. Again, no real surprises, which is not a bad thing — this hardware configuration performs very well on multiple fronts.

Sustained Write Performance and Cache Recovery

Official write specifications are only part of the performance picture. Most SSDs implement a write cache, which is a fast area of (usually) pseudo-SLC programmed flash that absorbs incoming data.  Sustained write speeds can suffer tremendously once the workload spills outside of the cache and into the "native" TLC or QLC flash. We use Iometer to hammer the SSD with sequential writes for 15 minutes to measure both the size of the write cache and performance after the cache is saturated. We also monitor cache recovery via multiple idle rounds.

The CS3140 writes at above 5.8 GBps for almost 20 seconds, revealing an SLC cache of around 116GB. This is not a particularly large cache but should be more than sufficient for a 1TB SSD. The drive then hits its direct-to-TLC phase at up to 2 GBps, which is quite fast for a 1TB drive. This reflects the choice of having a smaller SLC cache while using the faster 176-layer TLC. Performance would be higher at 2TB due to additional interleaving. As the cache is dynamic, it will shrink as the drive is filled. It’s still possible to hit a slower state with sufficient writing, but this is difficult to achieve.

Beyond flash choice, the size of the cache is one thing that separates drives based on the E18 controller. As we can see here with the Fury Renegade’s results, it has a larger, faster cache than the CS3140, but no middle-ground TLC state — this allows some flexibility for the manufacturer and the user. While cache recovery remained slow on both drives, the CS3140's ability to maintain TLC speeds ensures it has better steady state performance for edge cases.

Which is better is arguable, particularly as high-end PCIe 4.0 drives tend to focus on raw bandwidth. However, we believe in the future that sustained performance may become more important. For example, if you’re able to stream assets steadily from a drive via DirectStorage. That would be reads and not writes, of course, but there are nevertheless cases where consistency trumps peak performance alone.

Power Consumption and Temperature

We use the Quarch HD Programmable Power Module to gain a deeper understanding of power characteristics. Idle power consumption is an important aspect to consider, especially if you're looking for a laptop upgrade as even the best ultrabooks can have mediocre storage.

Some SSDs can consume watts of power at idle while better-suited ones sip just milliwatts. Average workload power consumption and max consumption are two other aspects of power consumption, but performance-per-watt is more important. A drive might consume more power during any given workload, but accomplishing a task faster allows the drive to drop into an idle state more quickly, ultimately saving energy.

We also monitor the drive’s temperature via the S.M.A.R.T. data and an IR thermometer to see when (or if) thermal throttling kicks in and how it impacts performance. Remember that results will vary based on the workload and ambient air temperature

Efficiency remains strong on the CS3140, thanks to the flash and the ability to quickly finish workloads with high peak performance. It still trails the Fury Renegade by a small bit.

According to both SMART and the IR gun, the CS3140 idles in the mid-40Cs and peaks in the upper-70Cs. This is hot but not unexpected as the drive lacks a heatsink. The smaller SLC cache helps when the drive hits the TLC flash, but sustained writes still get it pretty close to the throttling mark. In fact, we did see light throttling after about 215GB of writes when the drive hit 78C. Therefore, we recommend a heatsink if you anticipate having heavier workloads and/or a hotter environment.

Test Bench and Testing Notes

We use a Rocket Lake platform with most background applications such as indexing, windows updates, and anti-virus disabled in the OS to reduce run-to-run variability. Each SSD is prefilled to 50% capacity and tested as a secondary device. Unless noted, we use active cooling for all SSDs.

Conclusion

We didn't expect any surprises, and we didn't find any — the PNY XLR8 CS3140 gets close to our performance expectations without any weirdness. As with similar drives, the peak and general performance ranges from good to excellent. The SLC caching scheme is conservative, making for consistent performance in sustained workloads, and the new flash ensures good power efficiency. The PNY CS3140 is also a strong choice for a PlayStation 5 drive due to its option for a tailored heatsink.

On the other hand, there’s not much to set this drive apart from the competition. It is nice to see a functioning SSD toolbox — this is something lacking with, say, Inland drives — but the pricing isn’t particularly competitive. That being said, this is an adequate alternative and would be a good choice if you can find it on sale. PNY isn't pulling any endurance shenanigans here as far as we could tell, even if the company is a bit shy about some of the drive's specifications.

We scored the Kingston Fury Renegade a bit higher. The Fury Renegade edges out the CS3140 in almost every test while coming with cloning software, although currently, it is more expensive. The Kingston SSD is also designed for the PlayStation 5 and is perhaps a more trusted brand, perhaps except for those users who remember the ancient NAND switch-a-roo.

We also previously reviewed the Silicon Power XS70, a similar drive that's also designed for the PlayStation 5. It costs less than the CS3140 and comes with a PS5-capable heatsink out of the box. All of this makes it hard to give the PNY XLR8 CS3140 a higher score, but it is still a strong drive and contender. The biggest challenge for the XLR8 CS3140 is that it's in an increasingly crowded market, but it is a competent alternative.

MORE: Best SSDs

MORE: Best SSD for the Steam Deck

MORE: Best Hard Drives

MORE: Best External SSDs

PNY has quietly introduced its GeForce RTX 4070 Blower graphics card that, as the name suggests, uses a blower type cooling system. The board is currently available only in China, but it can be shipped to any part of the world and may eventually show up in the U.S. and/or Europe. 

PNY's GeForce RTX 4070 Blower, which was discovered by MEGAsizeGPU at JD.com, sports typical specifications for a GeForce RTX 4070 board: the AD104 graphics processor with 5,888 CUDA cores that operates at 1,920 MHz – 2,475 MHz and is connected to 12GB of GDDR6X memory using a 192-bit interface. The card has four display outputs (three DisplayPorts 1.4, one HDMI 2.0b) and one eight-pin auxiliary PCIe power connector.

The key selling point of the graphics card is not its enhanced performance, but its dimensions — it measures 278.4 mm x 119.4 mm  x 40 mm — that are compact enough to install more than one card into a motherboard thanks to the blower cooler.

Blower type coolers are not always the quietest or the most efficient cooling systems. However, when it comes to cooling down a graphics card that emits 200 ~ 300W of heat in a compact PC, these coolers are often as good as it gets. Blowers exhaust warm air from the case, efficiently operating even when internal air circulation is not ideal. Moreover, in cases where multiple graphics cards are combined in one system, these coolers are the only air solution that can ensure sufficient cooling in such confined spaces with limited air circulation. The only better option would be liquid cooling.

Nvidia doesn't often allow add-in-board partners to use blowers on consumer-oriented products. It's possible that this is because it allows multiple cards to be installed and more easily, and that that kind of horsepower that would rival that of professional graphics cards. Nvidia's RTX-branded professional cards are all equipped with blowers. 

But PNY's GeForce RTX 4070 blower is not cheap. The company officially calls the product 'artificial intelligence turbo graphics card' with a 'silent turbine fan suitable for multi-card environments' and positions it rather for AI developers than for gamers. To that end, the board is priced at ¥6,999 ($966 with taxes, $855 without taxes), which exceeds recommended price of a GeForce RTX 4070 in the U.S. ($599).

PNY had some of the best graphics cards on showcase at Computex 2023. The manufacturer even brought some workstation graphics cards, including the RTX 6000 Ada Generation and the RTX 4000 SFF. However, two particular graphics cards caught our attention.

First off, PNY has partnered with Cooler Master to produce the GeForce RTX 4090 Verto. It has become a trend for graphics card manufacturers to pair up with cooling specialists to create innovative products. For example, Asus taps into Noctua's experience and expertise to make the GeForce RTX 4080 16GB GDDR6X Noctua OC Edition.

If you think you've seen some thick GeForce RTX 4090 graphics cards, think again. For example, the GeForce RTX 4090 Verto is a 4.5-slot graphics card, which is insane because it covers almost all the PCI slots from your typical mATX case. For comparison, Asus' custom GeForce RTX 4080 with Noctua features a 4.3-slot design. The dimensions for the GeForce RTX 4090 Verto are 12.2 x 5.6 x 3.5 inches (310 x 142 x 89mm), so it's comparable to Asus' offering, which measures 12.2 x 5.7 x 3.5 inches (310 x 144.8 x 87.5mm). In PNY's case, Cooler Master designed the cooler, while PNY took care of everything underneath the shroud.

Not much is known about the GeForce RTX 4090 Verto. PNY didn't reveal the clock speeds for the GeForce RTX 4090 Verto. The shroud is a prototype, so that the retail version may look slightly different. The graphics card utilizes a dual-fan cooling solution and appears to feature a pair of Cooler Master's Mobius 120P ARGB 30th Anniversary Edition 120mm fans. The prototype has the cooling fans connected to the PWM fan connectors on the motherboard. It draws power from a standard 16-pin (12VHPWR) power connector and offers an HDMI 2.1 port and three DisplayPort 1.4a outputs.

There aren't many blower-type GeForce RTX 40-series graphics cards on the market. Thus far, we only know about two Ada Lovelace models that sport a blower design. In addition, Leadtek sells the WinFast RTX 4070 AI Blower 12G and the GeForce RTX 4070 Ti 12GB Hyper Brain, a China-exclusive SKU. Unfortunately, pickings are incredibly slim regarding blower graphics cards, and while Leadtek is a globally renowned brand, its products aren't widely available in the U.S. market. Therefore, PNY has an excellent opportunity for success with the GeForce RTX 4070 Blower Edition.

The GeForce RTX 4070 Blower Edition comes in a standard dual-slot format with dimensions of 11 x 4.7 x 1.6 inches (278.4 x 119.4 x 40mm), so it's a good fit inside small form factor (SFF) cases and workstations where you have multiple GeForce RTX 4070 Blower Edition graphics cards. The blower design is perfect for expelling heat outside the case instead of standard graphics card coolers. The GeForce RTX 4070 Blower Edition employs an 82mm fan for active cooling. Again, PNY didn't share the clock speeds for the graphics card.

The time has arrived — the 4060 Ti is now available for purchase. The reviews for the Nvidia RTX 4060 Ti cards are in, and you might be searching for a new card. Here we've compiled a list of the upcoming GeForce RTX 4060 Ti 8GB graphics cards in preparation for next week's launch, which will vie for a spot on our list of the best graphics cards. Best Buy remains Nvidia's exclusive partner for Founders Edition cards, and they'll be sold at the official $399 MSRP. In addition, below we have included several of the chipmaker's partners, including Asus, MSI, Gigabyte, Zotac, PNY, and Colorful. Some vendors' models have been listed early, while others will be listed in the coming days. Below we have the links to the models that are available at etail now, and we'll constantly update this article and add more links as the cards are listed over the coming days (we have the custom Gigabyte already listed, for instance). 

The list consists of GeForce 4060 Ti graphics cards featuring 8GB of memory since the 8GB model is the first to arrive on the market on May 24. However, 16GB models from AIB partners will be nearly identical to the ones shown here, with double the VRAM and a higher price tag. The exception is the Founders Edition model, which Nvidia will offer exclusively for the 8GB variant.

The GeForce RTX 4060 series was announced earlier, featuring the RTX 4060, RTX 4060 Ti 8GB, and RTX 4060 Ti 16GB. The RTX 4060 Ti 8GB highlighted in this article will launch on May 24 with a starting price of $399. Here's the rundown of all the models announced so far, how much they cost, and where to buy them.  

Nvidia

The GeForce RTX 4060 Ti Founders Edition will be the only 4060-series Founders Edition card. It will have 8GB of VRAM and looks to be nearly identical in design to the RTX 4070, except with a different color shroud. Under the hood, it will have a different PCB, GPU, and other components, but the aesthetics look good, and we expect it to set a reasonably high bar for different RTX 4060 Ti cards to clear.

Best Buy remains Nvidia's exclusive partner for Founders Edition cards, and they'll be sold at the official $399 MSRP. 

• $399.00 - Nvidia GeForce RTX 4060 Ti Founders Edition

Asus

Asus has revealed four custom GeForce RTX 4060 Ti graphics card and their respective overclocked variants. As usual, the ROG Strix GeForce RTX 4060 Ti model sits at the top of the product stack with the fastest clock speeds. In addition, the more compact  Dual GeForce RTX 4060 Ti comes in both black and white flavors.

• Asus ROG Strix GeForce RTX 4060 Ti 8GB GDDR6 OC Edition
• Asus ROG Strix GeForce RTX 4060 Ti 8GB GDDR6
• $459.99 - Asus TUF Gaming GeForce RTX 4060 Ti 8GB GDDR6 OC Edition
• Asus TUF Gaming GeForce RTX 4060 Ti 8GB GDDR6
• $399.99 - Asus Dual GeForce RTX 4060 Ti OC Edition 8GB GDDR6
• Asus Dual GeForce RTX 4060 Ti 8GB GDDR6
• Asus Dual GeForce RTX 4060 Ti White OC Edition 8GB GDDR6
• Asus Dual GeForce RTX 4060 Ti White 8GB GDDR6

MSI

MSI "wins" right now with a whopping eight different SKUs for the RTX 4060 Ti 8GB. Of course, that's four variants and then overclocked versions of each of those (the "Gaming X" cards are also overclocked, even though they don't include "OC" in their model names). We expect the Ventus cards to be the base models, though the triple-fan 3X may carry a slight price premium over the dual-fan 2X model. The Gaming models will have larger coolers and additional premium features.

• MSI GeForce RTX 4060 Ti Gaming 8G 
• MSI GeForce RTX 4060 Ti Gaming Trio 8G 
• $439.99 - MSI GeForce RTX 4060 Ti Gaming X 8G 
• $459.99 - MSI GeForce RTX 4060 Ti Gaming X Trio 8G 
• MSI GeForce RTX 4060 Ti Ventus 2X Black 8G 
• $399.99 - MSI GeForce RTX 4060 Ti Ventus 2X Black OC 8G 
• MSI GeForce RTX 4060 Ti Ventus 3X 8G 
• $419.99 - MSI GeForce RTX 4060 Ti Ventus 3X OC 8G

Gigabyte

Gigabyte has announced five different RTX 4060 Ti 8GB models so far, ranging from the compact Aero OC and the base model Eagle cards topping out with the Aorus Elite and Gaming OC models. There may also be non-OC variants of the Aero and Gaming cards, though those aren't currently listed. We expect the Eagle to sell for Nvidia's $399 MSRP, while the other four cards will likely carry a moderate price increase.

• $469.99 - Gigabyte Aorus GeForce RTX 4060 Ti Elite 8G
• $429.99 - Gigabyte GeForce RTX 4060 Ti Gaming OC 8G
• $449.99 - Gigabyte GeForce RTX 4060 Ti Aero OC 8G
• $419.99 - Gigabyte GeForce RTX 4060 Ti Eagle OC 8G
• $399.99 - Gigabyte GeForce RTX 4060 Ti Eagle 8G

Zotac

Zotac has announced three dual-fan RTX 4060 Ti 8GB graphics cards, including black and white versions. These cards have an aerodynamic (read: curved) aesthetic, a 2.2-slot wide cooler, and Zotac’s IceStorm 2.0 advanced cooling solution.

Zotac has also unveiled a Spider-Man: Across the Spider Verse-themed Twin Edge 4060 Ti 8GB model that comes with a bundle of themed goodies, including a tote bag and decorative decals to commemorate the movie that releases in theaters on June 2. In addition, the card has been modified with Spider-Man decals on the fans and a custom (magnetic) backplate featuring the same theme.

• Zotac Gaming GeForce RTX 4060 Ti Twin Edge 8GB 
• Zotac Gaming GeForce RTX 4060 Ti Twin Edge OC White Edition 8GB 
• Zotac Gaming GeForce RTX 4060 Ti Twin Edge OC 8GB - Spider-Man: Across the Spider-Verse Bundle

PNY

PNY also has three different 4060 Ti models. Two are base model configurations, meaning they should be priced at $399. The overclocked Verto card will likely have a modest price premium.

• PNY GeForce RTX 4060 Ti XLR8 Gaming Verto OC Edition 8GB 
• PNY GeForce RTX 4060 Ti XLR8 Gaming Verto Edition 8GB 
• $399.99 - PNY GeForce RTX 4060 Ti Verto Dual Fan Edition 8GB

Colorful

Colorful announced five different models, two of which come with a factory overclock. We don't usually see Colorful cards in the US as much as other brands, though the company sells many cards in Asia.

• Colorful iGame GeForce RTX 4060 Ti Advanced OC-V 8GB 
• Colorful iGame GeForce RTX 4060 Ti Ultra W-V 8GB 
• Colorful iGame GeForce RTX 4060 Ti Ultra W Duo OC-V 8GB 
• Colorful GeForce RTX 4060 Ti NB EX-V 8GB 
• Colorful GeForce RTX 4060 Ti NB EX Duo-V 8GB

It isn't uncommon for hardware makers and retailers to list yet-to-be-announced products weeks before they should officially start selling. It looks like this trend is repeating with Nvidia's GeForce RTX 4070. At least four vendors are already listing their RTX 4070-badged products, perhaps as a way to generate buzz and interest among consumers ahead of their official release, as noticed by @momomo_us and @harukaze5719.

MSI briefly listed its GeForce RTX 4070 Ventus 2X 12G graphics board on its website. The listing indicated that the product would carry 12GB of memory onboard and would feature a cooling system with two fans suggesting that we are dealing with a fairly compact graphics card, one of the first for the Ada Lovelace microarchitecture. Obviously, this is a preliminary listing, but since it is an official website, we can be more or less sure that it accurately reflects what MSI plans to offer. That said, take the news with a pinch of salt until the official release date arrives. 

Based on listings from a retailer, it seems that MSI also intends to release GeForce RTX 4070 Ventus 3X (with a bigger cooler), two Suprim-branded models, and two Gaming-badged boards aimed at enthusiasts and which are more than likely to enter the ranks of some of the best graphics cards available. So far, MSI and its partners seem to be offering at least eight GeForce RTX 4070 boards, which is a rather broad line-up. Prices of Geforce RTX 4070 offerings from MSI are likely to be significantly higher than the $599 allegedly recommended by Nvidia, though preliminary listed products tend to be overpriced in general.

Meanwhile, Gainward, Palit, and PNY also appear to prep multiple GeForce RTX 4070 versions, though these companies (or rather their partners) have yet to reveal more details about their offerings.

Listings of multiple GeForce RTX 4070 graphics cards weeks before launch could suggest that the product will be widely available right after the formal launch. Nvidia's GeForce RTX 4070 reportedly uses a cut-down version of the AD103 graphics processing unit with 5888 active CUDA cores, a chip that has been in mass production since late summer 2022. By now, the GPU developer should have plenty of silicon qualified for GeForce RTX 4070, so makers of graphics cards will have plenty of graphics chips to build products on.

The GeForce RTX 4090 continues to reign as the fastest of the best graphics cards currently available, and even better: prices seem to finally be dropping below $2,000. Take the PNY RTX 4090 XLR8 RGB, which you can pick up on Amazon — and there's the non-RGB variant for about $50 less. We're still hundreds of dollars above Nvidia's $1,599 starting MSRP, but we're getting closer. Maybe by the time the 40-series refresh cards start to show up we'll actually see acceptable prices (but don't count on it).

A big part of the problem with the RTX 4090 is the exceptional performance. It's up to 50% faster than the previous generation RTX 3090 Ti in gaming performance, without using DLSS 3. It's also significantly faster in many professional workloads as well as AI tasks. And the next step down, the RTX 4080, is still expensive and noticeably slower, so those with deep pockets are simply opting for the top product.

Three months on, after the newness has faded a bit and things are starting to settle down, how does the RTX 4090 look? We're looking at PNY's reference clocked RTX 4090 XLR8 Gaming Verto Epic-X RGB to find out — and kudos for that adjective laden product name, which we'll just truncate to "XLR8 RGB" for the rest of this review. 

It's a bit strange to have a second reference clocked card, as most companies tend to send us overclocked models. Obviously, that means identical specs to the RTX 4090 Founders Edition, and potentially lower performance than competing cards — but even heavily overclocked cards only add a few percent to framerates, often with higher power draw. So if you want reasonable power consumption and don't care about the final 3–5 percent increase in performance, this could be just what you're after.

More likely is that people will be interested in PNY's card if they can find one in stock for close to MSRP. All other areas being roughly equal, saving a few hundred dollars is always appreciated, even for extreme performance graphics cards.

As for the competition, AMD basically stops at the RTX 4080 level with its new RX 7900 XTX and XT GPUs. There's a relatively large gap between the 4090 and the next step down, whether you're looking at Nvidia or AMD cards. Despite the lack of Titan branding and certain Titan features (like unlocked drivers in a few professional applications), we effectively have a Titan replacement — at least until the actual future Titan RTX arrives (assuming it ever ships). 

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

PNY RTX 4090 XLR8 RGB Design

PNY has been around for decades as an Nvidia partner, and for a long time it only provided branded versions of the reference designs. That started changing around the GTX 10-series, when PNY created the XLR8 brand and started selling its own custom consumer models. Still, for most computer enthusiasts, PNY doesn't rank among the bigger names like Asus, Gigabyte, MSI (and formerly EVGA, RIP).

The packaging for the PNY 4090 box is fine, a bit more compact than other brands, but more than sufficient. The card itself is actually quite large, considering the box volume. However, the included support bracket is… weird. It's supposed to take the place of two of the motherboard screw mounts, but even if you can make that work, it only supports the back of the graphics card closest to the motherboard. GPU sag often occurs at the other side of the card, which hangs out in space, and this stand doesn't help fix that.

There's also the matter of branding. PNY's website refers to the card model as the "4090 XLR8 Gaming Verto Epic-X RGB," but the actual box omits any mention of "Verto" and the "Epic-X RGB" part of the name is a small logo. It's a small thing, but with three different PNY 4090 models shipping, it would be good to have the boxes clearly indicate which model is inside. But let's get to the card itself.

The PNY card measures 332 x 137 x 69 mm and weighs 1915g — not including the 4x 8-pin to 16-pin adapter. Of the five other 4090 cards we've looked at (Asus ROG Strix 4090, Colorful 4090 Vulcan, Gigabyte 4090 XLR8 RGB OC, MSI 4090 Suprim Liquid-X, and of course the 4090 Founders Edition), the only card that's physically smaller is the Founders Edition, and the PNY card weighs 75g less than the next closest Gigabyte card.

There's a full triple-slot IO bracket, which we appreciate as it provides a of extra stability. The card is technically a 3.5-slot design — so really it occupies four slots of space — which means you'll need a case with plenty of room. So far, there's not much to differentiate PNY's offering from other 4090 cards.

PNY uses a triple-fan setup with custom 98mm fans. There's an integrated rim and the blades appear slightly wider than we generally see on other cards. All three spin counterclockwise, but that's not a problem, as noise levels were very good.

Display outputs are the usual single HDMI 2.1 and triple DisplayPort 1.4a. While on paper the newer DP2.1 standard would be superior, in practice it doesn't matter much right now as DP2.1-capable monitors are practically non-existent. We're now using a Samsung Odyssey Neo G8 32 for our test system, which has 4K 240 Hz support that works fine with DP1.4, thanks to Display Stream Compression.

Power delivery comes via the usual 16-pin adapter that Nvidia supplies to all its partners. The uproar over melting adapters has died off now, which suggests Nvidia's claims that it was due to users not fully inserting the connector were correct. Whether the 16-pin connector is actually a good design is a different story, but it appears we'll be stuck with it now. 

PNY RTX 4090 XLR8 RGB Teardown 

There's not much reason for most users to take apart a graphics card, short of replacing fans or perhaps thermal pads if the default pads don't seem to be sufficient, but we like to get shots of the PCB to see if there's anything unusual. The disassembly process on the PNY card is pretty straightforward, though the components on the PCB are a bit interesting.

The first step is to remove the ten Phillips head screws securing the backplate. It can then be pulled off the card to expose the back of the PCB. Next, there are two more Phillips head screws under the backplate that secure the PCB to the cooler, plus the four screws (one with a sticker) that hold the GPU bracket in place.

You'll also need to remove the four Phillips screws that secure the IO bracket, as the bracket overlaps the PCB. That's 20 screws in total, which is about the norm for graphics cards of this size. Once those are out of the way, you can carefully pry the PCB, GPU, GDDR6X, etc. away from the cooler. But before you go too far, disconnect the three cables (two for fans, one for RGB lighting).

The cooler has a reinforcing frame that's secured by another four screws, but I didn't go any further. If you ever need to replace the fans, you'll want to focus on that part of the card and it shouldn't be too difficult — though finding the correct fans might be a different story.

As for the PCB, we've seen 4090 cards with anywhere from 20 to 26 phases for the GPU, plus another four phases for the GDDR6X. PNY's design only has 17 large black VRMs that are labeled LR22, plus three more for the GDDR6X. However, there are two additional L2R2 VRMs that may also be for the memory. So, this looks like a 17+3 phase design, but possibly 17+5 phases. That's less than other 4090 cards, but also this isn't an overclocked model and it's the first such card we've looked at.

The GPU is the expected AD102-300-A1 chip that's been used on all 4090 cards so far. It's possible Nvidia will have a revised chip at some point down the road, but this is a card from the 4090 launch. It just had to wait for the deluge of other GPUs before we could find time to finish the review.

Finally, the memory chips show "D8BZC," which corresponds to 16Gb chips rated for 21 Gbps. All of the 4090 cards have used memory with the same model, but half of the cards seem to have "downbinned" 24 Gbps chips that run cooler and can overclock to 24 Gbps or higher, while the other half typically max out at around 23 Gbps. Your results can and likely will vary based on your particular card, in other words. 

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

PNY RTX 4090 XLR8 RGB Overclocking 

Overclocking the PNY RTX 4090 was somewhat limited as the card is basically locked to a maximum of 450W. That's the theory, at least, but in practice, we were still able to increase the GPU core clock by +180 MHz and the memory reached a stable +1600 MHz (24.2 Gbps effective speed).

The good news is that average power use across our test suite only increased by about 3W, while performance at 4K was still 6.4% higher. That's a slightly lower gain than what we've seen on some of the other 4090 cards, but being able to do that without killing efficiency is a definite plus.

That brings up another point: If you're concerned about losing out on some performance by purchasing a reference card, manual overclocking can easily make up the difference. Yes, the fastest cards can also overclock and generally still end up a bit faster, but again with substantially higher power consumption. With the same power limit enforced, all the 4090 cards end up falling within a narrow range. 

PNY RTX 4090 Test Setup

We've recently updated our GPU test PC and gaming suite in late 2022, shifting from a 12900K and DDR4 memory to a 13900K with DDR5 memory, plus changing out one game and adding two new titles. The impetus for the change was that, with the RTX 40-series launch, we found more and more games were becoming CPU limited at anything below 4K.

The 13900K ranks as the fastest gaming CPU around (technically the 13900KS now slightly outperforms it), and we're still bumping into CPU limits at 1440p and especially at 1080p. But for third-party cards like this PNY 4090, we're skipping the 1080p testing.

Unfortunately, the five previous RTX 4090 cards were all tested on the old 12900K system, so we don't have comparable results for this review. We've retested the 4090 Founders Edition and it will stand in for the other GPUs. Note also that in our prior reviews, the fastest 4090 we've tested — the Asus RTX 4090 ROG Strix — was 1.2% faster overall in our rasterization tests and 2.1% faster overall in our ray tracing benchmarks. That's at 4K, and the benefit starts to drop off at 1440p. Anyway, keep those gains in mind as we dig into the PNY 4090 performance.

Along with the PNY card, we have results from the RTX 4070 Ti, RTX 4080, RX 7900 XTX/XT, and the previous generation RTX 3090 Ti and RX 6950 XT. We used the latest AMD and Nvidia drivers available at the time of testing: 22.12.2 for the 7900 series, 22.11.2 for older AMD GPUs; 527.62 for most of the Nvidia GPUs, and 528.02 for the PNY 4090. 

AMD and Nvidia both recommend either the AMD Ryzen 9 7950X or Intel Core i9-13900K to get the most out of their new graphics cards, and for the third party cards, we're going to stick with the 13900K. (The Ryzen 9 7950X provided similar performance on the reference 7900 cards.) MSI provided the Z790 DDR5 motherboard, G.Skill got the nod on memory, and Sabrent was good enough to send over a beefy 4TB SSD — which we promptly filled to about half its total capacity. The be quiet! 1500W PSU is ATX 3.0 compliant and 80 Plus Titanium certified.

We also have Nvidia PCAT v2 (Power Capture and Analysis Tool) hardware that allows us to capture the real-time power use, GPU clocks, and more during all of our gaming benchmarks. We'll have the results on the power testing page.

For all of our testing, we've run the latest Windows 11 updates. Our gaming tests now consist of a standard suite of nine games without ray tracing enabled (even if the game supports it), and a separate ray tracing suite of six games that all use multiple RT effects.

We've tested all the GPUs at 4K and 1440p, using "ultra" settings — basically the highest supported preset if there is one, and in some cases maxing out all the other settings for good measure (except for MSAA or super sampling). We've also hooked our test PCs up to the Samsung Odyssey Neo G8 32, one of the best gaming monitors around, so we could fully experience some of the higher frame rates that might be available — with G-Sync enabled.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

The RTX 4090 ranks as the fastest GPU currently available in our GPU benchmarks hierarchy, so we'll start with the 4K performance results.

Overall, the PNY card matches the Founders Edition, but with slightly different drivers we'd need to retest the FE to fully verify the results. Looking at the individual charts, PNY's card is about 2% slower in A Plague Tale: Requiem and Red Dead Redemption 2, while in Forza Horizon 5 it's 13% faster. Minimum FPS was also massively improved in Total War: Warhammer 3, which is interesting.

We suspect the latest 528.02 drivers corrected things in Forza and Warhammer, meaning in general performance probably ends up fractionally slower than the Founders Edition — not something you'd notice in normal use.

Overclocking is the usual story of modest gains, across the test suite. Overall performance was 4.7% higher, while in the individual games it was anywhere from 2.3% (Flight Simulator) to 7.8% (Warhammer 3) faster. Basically, the OC gains track with whether a game tends to be CPU limited, GPU limited, or GPU bandwidth limited, with Warhammer being an example of the latter.

Considering these gains come without increasing the power limit, at least overclocking in this fashion is a reasonable approach. We do appear to have "better" GDDR6X chips on this board, however, as we've noticed with half of the 4090 cards that they'll do over +1,500 MHz while the others will only do around +1,000 MHz, so these OC gains may not apply to every PNY board.

The ray tracing results are more in line with what we'd expect from the PNY card. PNY seems to have taken a somewhat more cautious approach with the VBIOS on this card, so it tries its best to avoid power spikes, even if that costs a bit of performance.

Overall, the PNY card is 2.9% slower than the Founders Edition, with individual results ranging from 0.3% slower (tied, in Spider-Man) to as much as 5% slower (Cyberpunk 2077 and Metro Exodus). Those last two dovetail nicely with the conjecture that power limits are keeping performance in check, as those are typically two of the most power hungry games we test.

Overclocking shows larger gains this time, to the tune of 8.2% overall, with a range of +2% (Spider-Man) to as much as 11% (Control and Metro).

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

Dropping the resolution to 1440p improves overall performance by 37%. That might seem like a lot, but it's also another indication that we're starting to bump into CPU limits — which is certainly the case with Flight Simulator. But consider the 3090 Ti, which sees a 57% increase in FPS by dropping the resolution, or the 6950 XT, which improves by 74%. In other words, the slower the GPU is, the less CPU bottlenecks come into play at 1440p.

PNY still ends up about 1% faster than the Founders Edition, but again that's likely drivers rather than actual performance differences — Forza Horizon 5 shows a 15% advantage for PNY now. We'll see about retesting the Founders Edition with the new drivers in Forza and Warhammer to see if that fixes things — and then we'll have to retest a few other RTX cards with the new drivers as well, most likely. It's a never-ending process.

Since ray tracing hits the GPU so hard, even at 1440p we still remain almost entirely GPU limited. We mentioned above that going from 4K to 1440p improved performance by just 37% on the PNY card in rasterization games, but in our DXR suite, the same resolution drop increased FPS by nearly 80%. The 3090 Ti improves by 89% and the 6950 by 104%, but at least in AMD's case it's probably more a case of the Infinity Cache hit rates dropping more at 4K.

The gap between the 4090 PNY and Founders Edition cards remains about 3% in DXR at 1440p, with minor differences across the test suite. PNY has taken a more conservative approach than even the reference card, which is somewhat unusual, but then the 4090 Founders Edition isn't really the typical reference design — if anything, it's overbuilt compared to previous generation reference cards.

We measure real-world power consumption using Powenetics testing hardware and software. We capture in-line GPU power consumption by collecting data while looping Metro Exodus (the original, not the enhanced version) and while running the FurMark stress test. We also check noise levels using an SPL meter. Our PC for these tests remains the same old Core i9-9900K as we've used previously, to keep results consistent.

We also test on our newer 13900K PC using PCAT v2 hardware in all of our gaming benchmarks, which gives a wider view of power use and efficiency. We'll start with the gallery of our PCAT results — note that FPS/$ uses the official MSRPs for the RX 7000- and RTX 40-series cards, while FPS/W (ie, efficiency) uses the measured power consumption.

We have results for our 1440p and 4K testing, along with the manually overclocked 4K testing (except on the RTX 4080 and 4090 Founders Edition cards, where we didn't retest overclocked results). We'll stick to just the latest generation GPUs for these tables.

The other GPUs are included mostly as a reference point, and our focus will be on the two 4090 cards. Interestingly, across our test suite, the PNY card ends up using slightly more power than the Founders Edition, at least for the overall geometric mean. There are some outliers in the results, though.

Look at Flight Simulator, where the Founders Edition only used 161W at 1440p while the PNY uses 381W — more than double the power use. Performance is basically unchanged, so it's possible that a driver update is to blame, or else just different VBIOS settings. But even omitting Flight Simulator, the PNY card consistently uses a bit more power at both 1440p and 4K, which likely comes down to differences in the board level components (capacitors, inductors, etc.)

The net result is that PNY's card ends up slightly less efficient overall, but it's not a big enough change to warrant further discussion. Overclocking meanwhile only increases power use by about 1% on average, as the power limit didn't change. A few games might use up to 2% more power, but the interesting bit is that clock speeds improve by 6% so overall efficiency also gets better.

Our previous power testing results are still valid, though they're limited to two scenarios: FurMark and Metro Exodus. Both test scenarios last about 10 minutes, though Metro does have a "break" between loops that allows the GPU to recover slightly.

As we've seen elsewhere, PNY's 4090 uses about the same amount of power as the Founders Edition card, coming in about 4W higher on average in Metro Exodus at the 4K extreme setting. In FurMark, the cards are basically tied on power use — no surprise since both exceed the official 450W TBP rating.

Clock speeds tell a similar story, with the Founders Edition averaging just slightly higher clocks in FurMark and Metro. But the real differences are in the temperatures and fan speeds.

PNY average 65C in Metro on the GPU, and 66C in FurMark. That's not a lot lower than the Founders Edition's 67C and 71C, but it's clearly better. That's especially true when you factor in fan speeds.

The 4090 Founders Edition already showed a good fan speed result of just 1409 RPM in Metro and 1505 RPM in FurMark, but PNY beats it in both cases. With FurMark, fan speed averaged 1359 RPM while Metro dropped to just 1228 RPM. That's about 37% fan speed, with plenty of room to go higher if needed — which is what we did with our overclocking tests, where fan speeds were in the 70% range.

Lower fan speeds mean less noise, and we test noise levels using an SPL (sound pressure level) meter placed 10cm from the card, with the mic aimed right at the center fan. This helps minimize the impact of other noise sources like the fans on the CPU cooler. The noise floor of our test environment and equipment is around 32 dB(A) with the PC powered on (the CPU cooler fans create a small amount of noise).

After running Metro for over 15 minutes, the PNY RTX 4090 settled in at a fan speed of 37% and a noise level of 38.2 dB(A). That's substantially better than the Founders Edition, which measured 45.0 dB(A) in the same test.

We also tested with a static fan speed of 75%, where the PNY generated 57.3 dB(A) of noise, but without overclocking or changing the fan speed curve, it's unlikely you'd ever see such levels. In fact, PNY seems to have tuned the card to stay below 50% fan speed based on our testing. 

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

Officially, the PNY 4090 XLR8 RGB is supposed to sell for $1,599, the same as Nvidia's Founders Edition. In practice, it's either out of stock or being sold by third-party resellers on places like Amazon and Newegg at $1,900 or more — just like the Founders Edition. It may be slightly less expensive than competing cards, so check around and see what's available if you're shopping for a 4090.

And that's really the major story here. If you're still trying to find an RTX 4090 card and you don't want to pay over $2,000, we've seen a few cards drop down as low as $1,700 — briefly — but supply still hasn't caught up with demand. And that demand isn't just from gamers, we'd wager.

The RTX 4090 might be the fastest card around for games, particularly if you want to play ray tracing games and you enable DLSS and/or DLSS 3. However, it's also equally performant in non-gaming workloads. In our initial review, the RTX 4090 in professional applications was around 80% faster than the RTX 3090 Ti in 3D rendering workloads.

That makes it a prime candidate for the "time is money" crowd, potentially cutting render times almost in half compared to the previous generation. The 4080 and 4070 Ti have since filled in some of the gap, but again, if you're waiting for 3D renders to complete and are willing to spend more money to get them done faster, the 4090 certainly fills that role as a prosumer card.

Sadly, even though GPU cryptocurrency mining is basically dead, it still feels like we're living in the world of 2021–2022 GPU prices. The new generation cards all cost more than their outgoing counterparts, and the cheapest RTX 40-series part right now, the 4070 Ti, costs more (in theory, anyway) than the old RTX 3080. Yes, it's 30% faster in our ray tracing suite, but MSRPs have gone up about 30% as well.

If you want the best graphics card, you'll need to decide what that means and how much you're willing to spend. For gamers, plunking down $1,600 for just the GPU feels outrageous, never mind spending $2,000 or more. Some people undoubtedly have that sort of pocket change available, but most gamers should look at the overall value instead.

And the best values are in the $300–$400 range, as usual. An RTX 3060 or RX 6700 XT might not win any bragging rights competitions, but you get over one-third the performance of the RTX 4090 for one-fifth the cost — that's according to our GPU benchmarks hierarchy. And if you're not aiming for 60+ fps at 4K, maybe you don't even need such an extreme GPU.

PNY doesn't break the mold when it comes to features and performance, but it does offer a solid product that runs cool and quiet — perfect for use in professional environments like an office space. It will also handle games basically as well as any other 4090, if that's your intent, and still keep noise levels down.

But as a value proposition? The RTX 4090 ranks dead last among modern GPUs right now. It would still be near the bottom of the stack even if you could buy it at MSRP, but at 25% over MSRP it's even worse. Maybe that will change in the future, but we wouldn't count on it.

• MORE: Best Graphics Cards
• MORE: GPU Benchmarks and Hierarchy
• MORE: All Graphics Content

PNY has created the Mako memory series under its XLR8 Gaming sub-brand. The XLR8 Gaming Mako family, which borrows inspiration from the shortfin mako shark, arrives in standard and RGB versions and comes in both DDR5-6000 and DDR5-6300 flavors.  

DDR5 pricing has improved significantly compared to when the new memory standard debuted. However, prices will continue to decline as we head into next year, so it's no surprise to see more value-oriented kits cropping up.

For now, the Mako series is the only enthusiast-grade DDR5 offering available from PNY, and it's only available at 32GB (2x16GB). In addition, the company has only listed DDR5-6000 and DDR5-6200 kits. However, the company's press release stated that the Mako would go up to DDR5-6400, so we may see another kit come to market soon. 

PNY's Mako memory modules sport an aluminum heat spreader that provides passive cooling to the integrated circuits (ICs) and power management IC (PMIC). The heat spreader only comes in black with a couple of diagonal silver lines to enhance the overall design. In addition, the company implanted the brand's XLR8 and Mako logos on both sides of the modules. 

The memory modules cater to builds where clearance space is a luxury, so the heat spreader is designed to be compact. As a result, the modules conform to a low-profile design with a height of 34.8mm (1.37 inches). For comparison, generic DDR5 memory measures 31.15mm (1.23 inches), so the Mako is only 11.7% taller.

Each Mako memory module has a capacity of 16GB; therefore, it features a single-rank design. PNY uses SK hynix H5CG48MEBDX014 (M-die) ICs for the memory module. Meanwhile, Richtek's "0D=8J 20H" PMIC handles the voltage regulation.

The memory posts at DDR5-4800 at 40-40-40-76 timings. It has an abundant amount of XMP 3.0 profiles. The main profile gets the memory to DDR5-6200, 42-42-42-88, and 1.3V. The alternate profiles configure the Mako to run at DDR5-5600, 40-40-40-77, and 1.2V and DDR5-4800, 38-38-38-77, and 1.1V.

See our PC Memory 101 feature and How to Shop for RAM story for more on timings and frequency considerations.

Comparison Hardware

Our DDR5 test system has Intel's Core i9-12900K Alder Lake processor with Corsair's CUE H100i Elite LCD liquid cooler taking care of the cooling. The 16-core Alder Lake flagship chip sits on the MSI MEG Z690 Unify-X motherboard, running the 7D28vA8 firmware. In addition, the MSI GeForce RTX 2080 Ti Gaming Trio is responsible for our gaming RAM benchmarks.

Our Windows 11 installation, benchmarking software, and games reside on Crucial's MX500 SSDs, whereas the RM650x feeds our entire system with the necessary power. Lastly, the Streacom BC1 open bench table ensures that our hardware is well-kept and tidy.

Intel Performance

The XLR8 Gaming Mako DDR5-6200 C42 lagged slightly behind the Viper Venom RGB DDR5-6200 C40 due to the latter's tighter memory timings. However, logically, the PNY memory kit delivered higher performance than the standard DDR5-4800 and DDR5-5200 memory kits.

Overclocking and Latency Tuning

The Mako memory uses SK hynix M-die ICs, similar to other DDR5-6200 memory kits. Therefore, it was easy to achieve a DDR5-6400 overclock at 1.4V. In addition, we could run the memory with 36-38-38-76 timings, the same configuration as Mako's rivals.

Lowest Stable Timings

When you run hardware outside of the manufacturer's specifications, there's always a potential risk of damage. That's the standard caveat with overclocking any hardware, not just memory. Although we've already reached out to SK hynix to inquire about the maximum safe voltage for its ICs, we haven't received any feedback on safe voltages for overclocking. However, after speaking with various memory vendors, they've agreed that 1.4V is the maximum voltage you would want to pump into DDR5 for an extended time.

Although PNY has opted for SK hynix M-die ICs, there's still a bit of silicon lottery regarding overclocking. For example, we could tighten the memory timings to 36-38-38-78 at 1.4V on the Mako, whereas the tRCD and tRP can go as low as 37 on the Viper Venom RGB DDR5-6200 C40 and Manta Xprism DDR5-6200 C36.

Bottom Line

The XLR8 Gaming Mako DDR5-6200 C42 prioritizes compatibility in the memory module's design and incorporates multiple XMP 3.0 profiles. While most high-speed DDR5 memory kits only come with a single profile, PNY's memory offers you two additional choices. Veteran users may find it redundant, but it's nice to have in case less experienced users happen to own a processor that can't do DDR5-6200. The out-of-the-box performance on the PNY XLR8 Gaming Mako DDR5-6200 C42 is average unless you manually optimize or overclock the memory.

The memory kit presently retails for $269.99 at multiple retailers. However, the price won't favor the memory and will scare off potential suitors. For example, the Vengeance RGB DDR5-6000 C36, which outperforms the XLR8 Gaming Mako DDR5-6200 C42, sells for $270. Alternatively, the Venom RGB DDR5-6200 C40, which is also faster, carries a $239.99 price tag, $30 cheaper than the PNY memory kit. 

If the pricing for the XLR8 Gaming Mako DDR5-6200 C42 improves, it could be a legitimate contender for anyone on the market for low-profile DDR5 memory or anyone who loathes RGB.

PNY has announced the new XLR8 Gaming Rev RGB series of DDR4 memory to compete with the best RAM on the market. In addition, the company promises aggressively overclocked memory with the top-tier ICs even though it's only offering the XLR8 Gaming Rev RGB in DDR4-3200 and DDR4-3600 flavors.

The memory targets the enthusiast segment, so the XLR8 Gaming Rev RGB arrives with the usual combination of an attractive, geometric heat spreader and dazzling RGB lighting. The latter is compatible with Asus Aura Sync, Gigabyte RGB Fusion 2.0, MSI Mystic Light Sync, and ASRock Polychrome Sync; therefore, you can customize the memory to your liking using your motherboard's includes software.

Although PNY speaks of top-tier ICs, it's still down to the lottery, which ICs you'll find in the XLR8 Gaming Rev RGB. The brand relies on all three big DRAM suppliers, including Micron, Samsung, and SK Hynix. Therefore, you wouldn't know what's under the hood until your purchase the memory.

PNY will sell the XLR8 Gaming Rev RGB as single modules (8GB or 16GB) and dual-channel memory kits (2x8GB or 2x16GB). The memory modules have XMP 2.0 support, PNY's limited lifetime warranty, and 24/7 U.S.-based technical support.

The DDR4-3600 memory has a CAS Latency (CL) of 18, whereas the DDR4-3200 variant has its CL configured to 16. Therefore, regardless of the data rate, both memory requires 1.35V.

The XLR8 Gaming Rev RGB is available today, and you can find the memory kits at your favorite retailer, such as Amazon and Best Buy, or directly from PNY.

Nvidia today released a new workstation graphics card called the RTX A4500. According to PNY's product page, this GPU comes with 7168 CUDA cores, 244 Tensor cores, 56 RT cores, and 20GB of GDDR6 with 640GB/s of memory bandwidth. This card is aimed at the workstation market and isn't intended to be a gaming GPU like Nvidia's standard GeForce RTX series GPUs.

The A4500 is designed to slide in between the current A4000 and A5000 GPUs released several months ago. The A4000 offers 6144 CUDA cores and 16GB of GDDR6 memory; meanwhile, the A5000 has a much higher core count of 8192 CUDA cores and 24GB of GDDR6 memory. For more info about these GPUs, check out our coverage here.

More specifications of the A4500 include ECC memory support, a PCI-e Gen 4 x16 interface, and four DisplayPort 1.4a ports. The card features NV-Link support as well for pairing multiple A4500s together for more intensive workloads. The card will operate with a maximum power output of 200W as well. According to PNY's model of the A4500, the card will also offer 23.7 TFLOPS of compute performance, which slots in between the A4000's 19.2 TFLOPS of performance and the A5000's 27.8.

The exterior of the A4500 features a matte black finish along with gold accents on the top and bottom sides of the card. The graphics card is two PCIe slots thick and features a blower-style cooler like many other workstation GPUs. Supplementary power is provided by a single 8-pin connector, allowing the card to reach its maximum power consumption of 200W.

Pricing for the A4500 is unknown since you have to order these GPUs from an OEM. However, according to Amazon prices, we expect it to hover around the $2,000 to $2,500 mark. The A4000 goes for $1,699 and the A5000 for $3,000 if you plan on buying this GPU from a retail source.

PNY has announced the company's performance DDR5 and XLR8 Gaming Mako memory kits, with the latter coming in both RGB and RGB-less flavors. The new memory kits will slot in just fine alongside Intel's 12th Generation Alder Lake processors.

The Performance series arrives with a bare PCB without any fancy heat spreaders or RGB lighting. The memory module is 16GB in size and comes in a single-DIMM presentation. It runs at DDR5-4800 (PC5-38400) with a CAS Latency (CL) of 40 and a 1.1V DRAM voltage. According to PNY, the Performance memory conforms to JEDEC's specifications, meaning that the memory module's other timings are probably 40-40-77.

The XRL8 Gaming Mako lineup, on the other hand, will sport a stylish aluminum heat spreader and optional RGB lighting. For those who opt for the Mako RGB, the heat spreader rocks an RGB diffuser comprised of a geometric polymer light pipe and super bright LEDs. The illumination is fully customizable through PNY's included software or your motherboard's lighting software.

PNY DDR5 Specifications

The XRL8 Gaming Mako memory kits are only available in a 32GB density with two 16GB DDR5 memory modules. In terms of data rate, PNY commercializes the XRL8 Gaming Mako series from DDR5-4800 up to DDR5-6400. In addition, the higher-binned offerings will have full support for XMP 3.0, facilitating the setup process and allowing you to define custom profiles.

The DDR5-5600 (PC5-44800) 32GB (2x16GB) memory kit is rated for CL36 with a 1.25V DRAM voltage. The DDR5-6500 (PC5-51200) is the flagship memory kit for the XRL8 Gaming Mako (RGB) series. The memory modules have a CL44 and require 1.4V to run.

PNY backs its Performance and XRL8 Gaming Mako (RGB) memory kits with a limited lifetime warranty. Like many vendors, PNY didn't share the pricing for the memory kits. So, unfortunately, they won't be up for purchase in the near future.

The Performance DDR5-4800 memory kits won't hit the shelves until mid-November, while the XLR8 Gaming Mako (RGB) will arrive in late Q4.

PNY has announced that its next-generation XLR8 Gaming DDR5-4800 memory kit will be ready for Intel's 12th Generation Alder Lake processors. While we've already seen Alder Lake chips with DDR5-4800 benchmarks, this is the first time that a company has confirmed that Alder Lake will arrive with DDR5 support.

The manufacturer stated in its press release that "with the Intel Alder Lake CPU launching late 2021, Intel’s CPUs will be the first to support the DDR5 standard, but with growing support from other CPU manufactures, DDR5 will likely overtake DDR4 in popularity by 2023."

As with any DDR5 memory, XLR8 Gaming DDR5-4800 memory modules will operate at 1.1V and come equipped with on-die single error correction (SEC) ECC support as well as built-in PMIC (Power Management IC) and a voltage regulator module (VRM) that aims to improve overclocking and voltage tweaking.

PNY didn't reveal the specifications for the memory, but did confirm that the memory modules will start at 16GB a piece. Assuming that it's a dual-channel memory kit, we're looking at a total capacity of 32GB (2x16GB). The CAS Latency (CL) could be 34, 40 or 42 clocks, depending on the bin.

Barring any mishaps, PNY expects to have its XLR8 Gaming DDR5-4800 memory available for compatibility testing in the third quarter of this year. Consequently, mass production will potentially commence in the fourth quarter. On the other hand, Alder Lake is rumored to arrive late 2021 or early 2022 so PNY's XLR8 Gaming DDR5-4800 memory should debut in time to accompany Intel's launch.

PNY unveiled its new LX family of SSDs that are designed specifically for write-intensive workloads, such as Chia plotting and mining. The company's LX-series drives deliver rather unprecedented endurance ratings of up to 54,000 terabytes written (TBW), courtesy to Phison's proprietary LifeXtension technology.

PNY's LX family includes three models that come in an M.2 2280 form-factor: the entry-level LX2030 2TB, the midrange LX3030 1TB, and the flagship LX3030 2TB. The drives feature a PCIe 3.0 x4 interface and are rated for up to 3200 MB/s sustained plotting read speeds and 2400 MB/s (LX3030) or 1000 MB/s (LX2030) sustained plotting write speeds. These aren't performance champions here, but the drives are tailored specifically for maximum endurance in plotting scenarios.

"I am very excited that PNY has partnered with Phison to deliver the first SSD for Chia plotting," said Jonmichael Hands, VP Storage Business Development at Chia. "The LX3030 is tuned for the Chia plotting workload, which requires a high amount of sustained bandwidth. Users can be at ease knowing they won't wear out the drive easily — the LX3030 1TB can create 2PB of plots before wearing out!" 

PNY LX-Series SSDs Specifications

The PNY LX SSDs are based on a Phison controller enhanced with the LifeXtension technology (designed specifically for Chia plotting), and presumably come coupled with 3D TLC NAND memory. Neither PNY nor Phison disclose how exactly LifeXtension works, but it looks like we are dealing with a compression technology designed specifically for Chia plots that contain hashes. They also indicate that it supports two XPlot modes: XPlot plain designed to increase endurance by about 3x and XPlot Pro that promises to boost endurance by up to 18x. 

It is unclear whether Phison's LifeXtension can be used for other write-intensive workloads to improve endurance of PNY's LX-series SSDs, but the company specifically stresses that these drives are designed for Chia plotting. 

Back in early May TeamGroup released its T-Create Expert PCIe SSDs, also tailored for Chia plotting and featuring high endurance ratings, but it looks like that drive has a lot of redundant 3D NAND memory for overprovisioning and does not use any technologies akin to Phison's LifeXtension.

PNY said that the LX-series SSDs are available immediately, but it did not disclose their recommended pricing. While the ability to create nearly 500,000 k32 Chia plots sounds impressive, storing those plots will be a costly endeavor. With Chia farming profitability currently sitting at around $2 per month for 1TB of storage, even with pooling, the long-term prospects aren't looking particularly promising.

PNY recently stepped into the spotlight when the company reduced the endurance of its XLR8 CS3030 SSDs, but apparently without fully informing its customers of the changes or the implications. The decrease in the warrantied write endurance, which was close to 80%, raised many questions - questions that PNY has answered in an official statement given to Tom's Hardware.

PNY doesn't deny having slashed the endurance on the XLR8 CS3030 lineup. The company attributed its decision to two factors: the rampant increase in demand for consumer SSDs for Chia coin farming, and the NAND shortage.

Although not one of the best SSDs on the market, the XLR8 CS3030 was one of the most durable. By giving the drive an endurance downgrade, PNY probably wants to dissuade Chia farmers from purchasing the drive.

PNY confirmed that it had evaluated other NAND options for the XLR8 CS3030 due to the global components shortage. While the performance remains intact on paper, the vendor updated the TBW (terabytes written) rating for the drives. The original XLR8 CS3030 used 3D TLC (triple-cell level) NAND. PNY didn't reveal any details on the substitute NAND. Given the lower TBW values, we suspect that PNY has switched over to QLC (quad-level cell) NAND, possibly explaining how the company now sells the XLR8 CS3030 with a 4TB model, a flavor that PNY didn't previously offer.

In fairness, the XLR8 CS3030's updated specification sheet doesn't commit to a certain SSD controller or NAND, leaving wiggle room for swapping components. PNY appears to have updated the document on May 17, 2021, so we assume that only units shipping after that date use different NAND. It's plausible that we have another SX8200 Pro situation where multiple variants of the same drive are floating around the market.

It would be almost impossible to differentiate one drive from another, at least by casual observation, since PNY uses the same part number. At any rate, the vendor believes that the warranty period will, in all likelihood, expire before an average consumer hits the TBW threshold. Therefore, the product's lower endurance shouldn't affect consumers that "use the SSD as intended."

Lastly, PNY addressed how the company will handle the warranty on the XLR8 CS3030. The limited five-year warranty is still valid. However, drives sold before May 17, 2021, abide by the previous TBW ratings, while SSDs sold after that date are covered under the new specifications.

You can find PNY's complete statement below:

PNY’s Official statement: 

The changes PNY made to its XLR8 CS3030 SSD’s warranty policy were driven by two factors, the uptick in demand for using high-speed, consumer-grade SSDs for Chia farming, and the industry-wide shortage of NAND. These changes were published and made public on the company’s website in both the warranty section as well as the CS3030 product spec sheet on May 17, 2021.

Why TBW was added to PNY’s CS3030 SSD warranty: 

The onset of Chia farming has many PC component brands rethinking their warranties, as consumer-grade hardware is not typically under the type of intense write use that is synonymous with Chia farming. The write activity required to farm Chia coin can wear out typical consumer-grade SSDs in a matter of weeks. Because of this, PNY, like others, introduced a Terabytes Written (TBW) policy to its SSD warranty. For consumers using these SSDs as intended, the warranty time (years) period will likely run out before they hit the TBW thresholds.
Source: https://www.tomshardware.com/news/galax-kills-warranty-ssd-mining. 

Why the CS3030 TBW rating was reduced: 

Due to an industry-wide shortage of NAND, PNY qualified additional NAND options for the XLR8 CS3030 SSD. While the read/write performances met or exceeded published specs, some of the TBW endurance was lower thus PNY set its warranty threshold and updated the sell sheet based on the lowest TBW rating of those qualified options. For consumers using these SSDs as intended, the warranty time (years) period will likely run out before they hit the TBW thresholds. 

PNY’s SSD warranty coverage:

Drives sold prior to May 17, 2021, correspond to the previously posted warranty, whereas drives sold on May 17, 2021, and later correspond to the latest warranty and TBW thresholds. Again, most consumers that use these drives as intended will likely exceed the warranty time (years) period before crossing the TBW threshold.

ComputerBase, a German news outlet, reported that PNY has significantly reduced the endurance rating for the brand's XLR8 CS3030 M.2 NVMe SSD. The recent change suggests that PNY might have secretly switched out some of the drive's core components for lesser quality parts.

The XLR8 CS3030 is a M.2 2280 SSD that runs on the conventional PCIe 3.0 x4 interface. The drive offers sequential read and write speeds up to 3,500 MBps and 3,100 MBps, respectively. Apparently, the performance between the new and previous revisions remains unchanged. While comparing the old datasheet with the new datasheet, we did notice that PNY no longer specifies the type of NAND that's used inside the XLR8 CS3030. For reference, the original XLR8 CS3030 used TLC (triple-level cell) chips. Although at this point in time, it's unknown if PNY replaced the NAND, the SSD controller or both.

It's not unusual for SSD manufacturers to swap out components. Big-name vendors, such as Kingston, Adata and Crucial, have all done so in the past. The problem arises when the maker doesn't communicate the changes to the consumer. Not every manufacturer produces their own components, so companies typically buy what's available on the market. Factors such as product cycles or shortages play a big role in whether vendors can secure the necessary components that were used in the original product.

PNY XLR8 CS3030 Specifications

At least on paper, the XRL8 CS3030's performance seems unaffected. However, it'll require thorough testing to see whether the new revision performs equally as the previous revision. In PNY's case, the latter revisions of the XLR8 CS3030 arrive with a substantial reduction in endurance, which can be just as important as performance.

The 500GB, 1TB and 2TB variants of the XLR8 CS3030 all exhibit up to 78.8% lower TBW (terabytes written) ratings — just a few figures shy of 80%. Even the entry-level 250GB version suffers from a 55.3% reduction. Despite the lower lifespan, PNY still lists the XLR8 CS3030 with a limited five-year warranty, same as before.

SSD warranties work in two ways. The manufacturer backs the drive for a certain time period or a rated endurance, whichever comes first. While PNY maintained the five-year warranty, the XLR8 CS3030 SSDs have lower endurance rating. If you're the type of consumer that torture your SSDs, the drive will likely exceed its endurance value before the five years is up.

The biggest question that PNY has to answer is whether the company will honor the previous or new TBW ratings for drives that are on the market. More importantly, the brand will have to define how it will handle with the warranty on the XLR8 CS3030 for consumers who've already purchased the drive. The specification sheets show that both the original and new revisions carry the same part number so PNY isn't openly disclosing the change.

For now, it's anyone's guess to why PNY modified the life expectancy of its XLR8 CS3030 SSDs. With the recent hype with farming Chia coin, PNY might just be frightened to have to deal with RMAs. Farming Chia on consumer drives is a bad idea since they'll probably perish in a matter of weeks, which is why Chia-designed SSDs like Sabrent's Plotripper exist.

ComputerBase reached out to PNY, and the company has admitted to the alteration. However, PNY refused to comment on what has change or the reason for the change. Nevertheless, we've reached out to PNY to see whether the brand will provide any further insights on the situation.

Nowadays, it feels like the norm that every computer hardware company has a dedicated gaming sub-brand. For PNY, that would be XLR8 Gaming that currently competes in three major hardware markets: memory, gaming graphics cards, and SSDs. In terms of memory, the XLR8 Gaming branding is still a bit wet behind the ears, but the company has started to solidify its lineups. The Epic-X RGB series, in particular, is one of XLR8 Gaming's latest additions to its memory portfolio.

The Epic-X RGB memory modules come with a black PCB and a matching aluminum heat spreader. The design is as simple as it gets, and that's not a bad thing. The heat spreaders feature a few diagonal lines and the XLR8 logo in the middle. An RGB lightbar is positioned on top of the memory module to provide some flair. The memory measures 47mm (1.85 inches) tall, so it might get in the way of some CPU air coolers.

PNY didn't develop a proprietary program to control the Epic-X RGB's lighting, which will favor users who don't want to install another piece of software on their system. Instead, PNY is handing the responsibility over to the motherboard. Fear not, because the Epic-X RGB has all its bases covered. The memory's illumination is compatible with Asus Aura Sync, Gigabyte RGB Fusion, MSI Mystic Light Sync, and ASRock Polychrome Sync.

The Epic-X RGB memory kit is comprised of two 8GB DDR4 memory modules. They're built on a 10-layer PCB and feature a single-rank design. Thaiphoon Burner was unable to identify the integrated circuits (ICs) inside the Epic-X RGB. However, given the primary timings, the memory is likely using Hynix C-die chips.

Predictably, the Epic-X RGB runs at DDR4-2133 with 15-15-15-36 timings by default. There's a single XMP profile that brings the memory up to speed. In this case, it sets the memory modules to DDR4-3200 and the timings to 16-18-18-38. At this frequency, the memory draws 1.35V. For more on timings and frequency considerations, see our PC Memory 101 feature, as well as our How to Shop for RAM story.

Comparison Hardware

Our Intel test system consists of an Intel Core i9-10900K and Asus ROG Maximus XII Apex on 0901 firmware. On the opposite side, the AMD testbed leverages an AMD Ryzen 5 3600 and ASRock B550 Taichi with 1.30 firmware. The MSI GeForce RTX 2080 Ti Gaming Trio handles the graphical duties on both platforms.

Intel Performance

Predictably, the Epic-X RGB didn't beat the faster memory kits in our RAM benchmarks. Performance was consistent, with the Epic-X kit placing third overall on the application and gaming charts.

AMD Performance

Things didn't change on the AMD platform, either. However, the Epic-X RGB did earn some merits since the memory kit was the fastest in the Cinebench R20 and HandBrake x264 conversion tests. The margin of victory was slim, though, at less than 1%.

Overclocking and Latency Tuning

The Epic-X RGB isn't the best overclocker that we've had in the labs. Nevertheless, we squeezed an extra 400 MHz out of the kit. We could hit DDR4-3600 at 1.45V after we relaxed the timings to 20-20-20-40.

Lowest Stable Timings

Sadly, we didn't have the same level of luck optimizing the Epic-X RGB at DDR4-3200. Even with a 1.45V DRAM voltage, we could only get the CAS Latency down from 15 to 16 clocks. The other timings wouldn't yield.

Bottom Line

In this day and age, enthusiasts are pursuing faster and faster memory kits. However, there's always space for a standard memory kit, and the XLR8 Gaming Epic-X RGB DDR4-3200 C16 kit could very well find its place with users that want to stick to a processor's official supported memory frequency. Today's modern processors, such as AMD's Zen 2 and Zen 3 processors and Intel's looming Rocket Lake processors, support DDR4-3200 memory right out of the box. The XLR8 Gaming Epic-X RGB DDR4-3200 C16 would fit nicely in this situation since you can just enable XMP and never look back.

The XLR8 Gaming Epic-X RGB DDR4-3200 C16 only has a tiny flaw, and that's pricing. The memory kit retails for $94.99 when the typical DDR4-3200 C16 kit starts at $74.99. Even the faster DDR4-3600 C18 memory kits sell for as low as $79.99. In PNY's defense, the Epic-X RGB memory modules do look nice with the RGB lighting and whatnot, so we can probably chalk the extra cost up to the RGB tax.

Not a lot of companies have announced Mini-ITX versions of their Nvidia GeForce RTX 3060-based graphics cards so far, but among them is PNY, one of Nvidia's premier partners. One store in Germany claims to have stock of PNY's GeForce RTX 3060 board designed for compact machines, and there are three things that immediately attract attention: a minimalist design, integrated addressable RGB lights, and a very long name. 

From a specifications point of view, PNY's GeForce RTX 3060 12GB XLR8 Gaming Revel Epic-X RGB Single Fan Edition is a rather regular Nvidia GA106-300-based graphics card with 12GB of GDDR6 memory. The board has a stock 1320 MHz clock, which boosts up to 1807 MHz, slightly higher than a stock RTX 3060 12GB. It carries a single-fan dual-slot cooling system, requires an eight-pin auxiliary PCIe power plug, and features four display connectors: three DisplayPort 1.4a and a single HDMI 2.1 port.

The card looks as minimalistic as some of its Mini-ITX counterparts from other manufacturers, but since it has integrated Epic-X RGB lighting, it immediately attracts the attention of enthusiasts and modders. Not all components aimed at Mini-ITX builds feature RGB LEDs, so for this PNY deserves praise. 

PNY formally added its Mini-ITX GeForce RTX 3060 graphics card to the lineup a few weeks ago. Meanwhile, it is already listed at Amazon.NL with availability date of February 25, 2021. The board is not currently available, but the listing may indicate that the online store expects it to hit its virtual shelves shortly. Unfortunately, Amazon does not publish any information about pricing of the graphics card. 

Only a few hours are left before the GeForce RTX 3060 goes up for purchase at retailers around the world. Given Nvidia's recent launch history, the mid-range Ampere graphics card's stock might be equally limited so legit customers will likely have to fend off cryptocurrency miners and scalpers once again.

If you haven't been following the GeForce RTX 3060, here's quick summary. Built with the GA106 silicon, the GeForce RTX 3060 comes equipped with 3,584 CUDA cores, 112 Tensor cores and 28 RT cores. The reference clock speeds include a 1,320 MHz base clock and 1,777 MHz boost clock, but the more expensive custom models will likely exceed those figures.

Designed to compete with the best graphics cards in its price band, the GeForce RTX 3060 has a generous 12GB of 15 Gbps GDDR6 memory at its disposal. With a 192-bit memory interface, the graphics card delivers a maximum memory bandwidth up to 360 GBps. While its real-world output remains to be seen, the GeForce RTX 3060 could offer performance that's between last generation's GeForce RTX 2070 and RTX 2070 Super.

Nvidia GeForce RTX 3060 Pricing

Best Buy and Newegg are the first retailers to put up the GeForce RTX 3060. While Newegg has seemingly removed the pricing, hardware detective momomo_us took a screenshot of the original listings. Since the launch window is so close, the prices quoted should be the final pricing.

Thus far, PNY's GeForce RTX 3060 12GB XLR8 Gaming Revel Epic-X RGB Dual Fan and Single Fan are listed at Best Buy for $629.99 and $624.99, respectively. It's uncertain if there was a pricing error or Best Buy really intends to sell the PNY models for those ridiculous prices.

Don't expect to find the high-end models, such as the MSI GeForce RTX 3060 Gaming X Trio 12G, Gigabyte Aorus GeForce RTX 3060 Elite 12G or Asus ROG Strix GeForce RTX 3060 OC for anything less than $500 though. For comparison, you could end up paying up to 64.1% more than Nvidia's $329 MSRP. Of course, these premium models will come with all the bells and whistles.

Ultimately, budget buyers will find refuge in either Gigabyte's GeForce RTX 3060 Eagle or EVGA's GeForce RTX 3060 XC Black Gaming. These two models are the only ones that retail for $329.99. EVGA hasn't activated the product pages for its GeForce RTX 3060 offerings yet. As far as the GeForce RTX 3060 Eagle goes, Gigabyte listed it with a 1,777 MHz boost clock so it's basically running within Nvidia's reference specification.

Big-time miner Simon Byrne went the extra mile and made a portable RTX 3080 mining rig that is capable of going 0-60mph in under 4.5 seconds. 

The vehicle Byrne used is the BMW i8, and in the trunk is a small RTX 3080 farm with six GPUs. Simon uses this machine to make money on the go when he is away from his massive 78 RTX 3080 mining farm.

We reached out to Byrne to verify this system is real, and he confirms that he made it himself. In fact, when we asked if it was real, he responded: "Indeed, just to annoy gamers."  

Byrne also gave us some additional info on how well the farm functions inside the car. As it turns out, the BMW i8 has more than enough power from its battery pack to power the miniature RTX 3080 farm. The car can output up to 3500W from its main outlet, and the mining rig consumes just 1500W. The only "issue" is that the rear hatch/trunk needs to stay open, or the mining rig will overheat. (He also says that the system is connected using an inverter.)

If you are unfamiliar with the BMW i8, the hybrid car is optimized towards performance with 369 horsepower in total, but just 141hp comes from the i8's combustion engine, which means more than half of the car's power output comes from electric motors. To cope with all that power, the i8 comes with an 11.6kWh 353V lithium-ion battery pack; in layman's terms, the i8's battery pack is more than adequate to power several gaming/mining rigs if necessary.

Byrne also shared his broader opinion of cryptocurrencies, saying, "Cryptocurrency is open-sourced & decentralized. It's trustless, and there isn't interference from a third party. No one owns or controls it, and it does not have a governing body."

"Crypto may be driven by code, but it's fueled by a passionate, altruistic, open-sourced community. Individuals who voluntarily invest their time & energy helping others without coercion or the threat of violence. They recognize that their success is dependant on the success of others. I'm not in Crypto to make a fast buck, but I'll take the bonus; I'm in it because it's a vision of a utopian world where services are acquired through voluntary relationships."

Tomorrow's RTX 3070 launch is fast approaching, and in preparation, Best Buy has begun to reveal RTX 3070 listings in its online store. So far they have released pricing and product information on the Nvidia RTX 3070 Founders Edition, Gigabyte RTX 3070 Eagle, Gigabyte RTX 3070 Gaming, and two PNY RTX 3070 models.

The Founders Edition will cost $499.99 and is the cheapest out of the five right now, it features a compact dual-fan design with a similar cooler to the RTX 3080 Founders Edition. The card is pretty small, measuring just 9.5" in length, so it should fit in a lot of compact cases. It also comes with a single 8-pin power connector with a 12 pin adapter if required. 

Next is the Gigabyte RTX 3070 Eagle, which will also cost $499.99. The Eagle brand of products is Gigabyte's new lineup for Ampere, aiming to be a new budget-oriented option for buyers. The card includes a triple-fan cooler measuring 2 slots in thickness. It's not as fancy as the higher trims, but it'll do its job well in cooling the 3070 GPU. 

Gigabyte also has the slightly more advanced RTX 3070 Gaming, and that will be $569.99. The Gaming brand is Gigabyte's mid-range product with a bit more cooling power vs. the Eagle. It's also a triple-fan cooler but slightly thicker at 2.5 slots and features a slick black and silver theme. One feature setting this card apart is its dual-BIOS capability. One BIOS provides a factory OC for better performance and the second runs on a silent mode, presumably with a slower fan curve for quiet operation.

Up next is the PNY RTX 3070 XLR8 Gaming Epic-X, coming in at $539.99. This is PNY's top trim RTX 3070 right now. It also has a triple-fan cooler design, which means it's pretty large at 11.57 inches in length. Fortunately, it's not taller than the PCIe brackets, so compatibility in that area is strong. The front features rings of RGB lighting around the center fan, with metal bars around all three fans which serve as wire guards to help protect the blades.

Best Buy also lists PNY's more compact 3070, the PNY RTX 3070 Dual fan. This card (like its namesake) is a compact dual fan design resembling the dimensions of the 3070 Founders Edition. It is slightly bigger though, at almost 3 slots in thickness. Pricing is the same as the XLR8 at $539.99.

Best Buy's prices should hopefully give you some baseline expectations as to where the rest of the RTX 3070 AIB partner cards will land. It seems the cheapest cards, like the Eagle, will cost the same as the Founders Edition, while mid-range models like the Gaming will get a slight uptick in price to the mid-$550 range. The Founders Edition looks to be the best card in terms of price and form factor, as it's the smallest of the five (including the PNY Dual Fan). But if it's anything like the other FE Ampere cards, availability will probably be slim. Striking is that both PNY cards share the same $539.99 price. With how much larger the XLR8 is, I'd expect it to cost more.

The RTX 3070 Founders Edition and its AIB partners officially release tomorrow, so expect full details on 3070 prices then. For a full list of RTX 3070 models as we find them, check our guide on where and how to buy RTX 3070 cards.

The GeForce RTX 3070 officially goes on sale on October 15, but Spanish retailer PcComponentes (spotted via overclocking.com) has already posted a couple of custom models on its online store.

Armed with Nvidia's Ampere silicon, the GeForce RTX 3070 aims to become the best graphics card in the budget category. Its attributes include 5,888 CUDA cores, 46 RT cores and 184 Tensor cores. The 8GB of onboard 14 Gbps GDDR6 memory runs through a 256-bit memory bus to offer a memory bandwidth that peaks at 448 GBps. 

The GeForce RTX 3070 is rated for 220W and utilizes the 12-pin PCIe power connector. The Founders Edition will retail for $499, but the design might not appeal to everyone, which is where Nvidia's partners come in with their custom models.

Custom GeForce RTX 3070 Prices

Based in Spain, PcComponentes deals in Euros, so we've converted the prices over to U.S. dollar sand deducted the standard 21% VAT (value-added tax). But keep in mind that computer hardware overseas are rarely priced the same as in the U.S. PcComponentes' postings could serve as an estimate but not definitive pricing we should expect in the U.S.

The ROG Strix GeForce RTX 3070 OC Edition, which could retail for about $653, is the most expensive custom SKU in PcComponentes catalogue. Asus hasn't expose the clock speeds for the triple-fan monster. For reference, the Founders Edition's boost clock is set at 1,725 MHz. Asus' non-OC edition has a boost clock of 1,755 MHz so we expect the OC variant's boost clock speed to surpass 1,800 MHz. 

Assuming that the pricing is accurate, Zotac's Gaming GeForce RTX 3070 Twin Edge could be the cheapest custom option. The dual-fan graphics card has the same boost clock as the Founders Edition.

The RTX 3070 game begins October 15. If the GeForce RTX 3070's stock is anything like the GeForce RTX 3080 or RTX 3090, things will be hectic with genuine consumers battling scalpers and bots to secure a GeForce RTX 3070.

The GeForce RTX 3070, which could be a serious contender for the best graphics card in the mid-range category, lands on October 15. Despite that, Amazon has already listed PNY's GeForce RTX 3070 8GB XLR8 Gaming Epic-X RGB Triple Fan Edition and GeForce RTX 3070 8GB Dual Fan, as spotted via @momomo_us. 

Despite being custom models, PNY's iterations respect Nvidia's reference specifications for the RTX 3070. Built on Samsung's 8N Nvidia process node, the graphics card arms itself with 5,888 CUDA cores, 46 RT cores and 184 Tensor cores. It operates with a 1,500 MHz base clock and 1,725 MHz boost clock. The 8GB of 14 Gbps GDDR6 memory runs across a 256-bit memory interface for a theoretical maximum throughput up to 448 GBps.

Staying true to their model names, PNY's GeForce RTX 3070 8GB XLR8 Gaming Epic-X RGB Triple Fan Edition and GeForce RTX 3070 8GB Dual Fan utilize cooling solutions featuring dual-fan and triple-fan setups, respectively.

The RTX 3070 8GB XLR8 Gaming Epic-X RGB Triple Fan Edition is the more gaudy model, boasting an aggressive shroud and RGB lighting. The graphics card adheres to a 2.7-inch design though and measures 293.88mm (11.57 inches) long. 

Contrastingly, the GeForce RTX 3070 8GB Dual Fan sticks to a more minimalist look that drops the flashy lighting. There aren't dimensions listed for this graphics card, but we expect it to be substantially shorter than the top model.

PNY made differing decisions on the PCIe power connectors. The GeForce RTX 3070 8GB XLR8 Gaming Epic-X RGB Triple Fan Edition utilizes two conventional 8-pin PCIe power connectors, while the GeForce RTX 3070 8GB Dual Fan jumps aboard Nvidia's 12-pin PCIe power connector bandwagon. At the end of the day, it doesn't really matter which PCIe power connectors are present on the graphics cards, since both are rated for 220W.

There aren't any surprises with the video outputs. The pair of PNY graphics cards come equipped with one HDMI 2.1 port and three DisplayPort 1.4 outputs, which is the same setup on Nvidia' GeForce RTX 3070 Founders Edition.

Speaking of the Founders Edition, Nvidia sells it for $499. PNY's custom models will likely carry a small premium over the Founders Edition. 

Nvidia has disabled preorders with Ampere, meaning everyone gets a fair chance at snatching up a GeForce RTX 3070 come October 15.

Nvidia's RTX 3080 and 3090 GPUs are launching soon, with the RTX 3090 and RTX 3080 founders edition cards coming out later this month. But don't forget about aftermarket cards from Nvidia's AIB partners. A recent twitter post from @momomo_us revealed full pricing for RTX 3080 and RTX 3090 aftermarket cards from Caseking.de, a German retailer site.

The retailer lists full pricing on several SKUs from Gigabyte, Asus, Zotac INNO3D, PNY and Asus. Only some SKUs from each manufacturer were listed, with the exception of INNO3D, which has its entire lineup listed.

Here's a look at pricing for each aftermarket card in the listing. For reference in Europe, the RTX 3080 Founders Edition costs $710 and $1531 for the RTX 3090.

All prices have been converted from German to USD currency:

The pricing strategy is similar to the last generation cards. Cheaper base model aftermarket graphics cards sit closer to the founders edition pricing, while more exotic offerings like the STRIX and Trinity sit around $100 more than FE pricing. It's interesting to see none of these cards are priced below that the Nvidia Founders Edition cards. For this generation, aftermarket cards are much larger and cooling has been beefed up tremendously, so it makes sense that prices have increased.

There's still no word yet on when AIB partner cards are slated for release. All we know for now is the RTX 3080 will launch September 17th and the RTX 3090 on September 24th. Both are founders edition cards. Hopefully, aftermarket cards will launch in tandem, or soon afterward.

If you're looking to build a system with a ton of bling, PNY's latest XLR8 EPIC-X RGB memory kits might be just what you're looking for. These are the company's latest announcement, expanding on the XLR8 lineup with an RGB option as spotted by TechPowerUp.

The XLR8 EPIC-X RGB DDR4 modules will come in four different kits at launch, featuring 8 GB, 16 GB with two 8 GB modules, 16 GB, and 32 GB with two 16 GB modules. All the kits run at speeds of 3200 MHz with CL 16-18-18-36 timings, and they'll require 1.35V to operate. So, as far as specifications go, we're looking at some pretty ordinary mid-tier level memory, but most people won't observe performance gains from faster memory anyway.

The coolers installed onto the XLR8 EPIC-X RGB memory modules are made of aluminum alloy, and topped off with diffusers for a smooth, soft lighting effect provided by the LEDs underneath. The lighting is provided by five RGB LEDs underneath each diffuser.

The RGB lighting on the modules is compatible with the Asus Aura Sync, Gigabyte RGB Fusion, MSI Mystic Light Sync and ASRock Polychrome Sync ecosystems.

The new XLR8 EPIC-X RGB memory kits are available for purchase on PNY's website. The 8GB and 16GB modules cost $39.99 and $69.99, respectively. The 16GB (2x8GB) and 32GB (2x16GB) memory kits go for $79.99 and $134.99, respectively.

There are many 512GB USB flash drives on the market, but very few come in a small presentation like PNY's Elite-X Fit offerings. As per a ComputerBase report, the Elite-X Fit is now available with a 512GB capacity.

The Elite-X Fit measures 2.16 x 1.52 x 1.55 cm, so it isn't a lot longer than the typical USB Type-A connector itself. The drive adheres to the USB 3.1 Gen 1 specification and, according to PNY, the Elite-X Fit 512GB (P-FDI512EXFIT-GE) delivers read speeds up to 200 MBps. PNY rates write performance as 30 times faster than USB 2.0 (3 MBps), which works out to 90 MBps.

Given the Elite-X Fit's dimensions, the SanDisk Ultra Fit is the only competitor within the same size class. The Ultra Fit, however, comes with a 2.97 x 1.42 x 0.51cm footprint, so the Elite-X Fit is a tad shorter. 

SanDisk advertises the Ultra Fit 512GB with a read speed up to 130 MBps. However, the manufacturer doesn't commit to a write speed, instead just touting performance up to 15 times faster than a normal USB 2.0 flash drive. By SanDisk's standards, USB 2.0 flash drives write up to 4 MBps, which would put the Ultra Fit's write performance to around 60 MBps. In summary, PNY's drive reads 53.8% and writes 50% faster than SanDisk's drive.

PNY sells the Elite-X Fit 512GB for $99.99 on the brand's online store. The Ultra Fit 512GB typically costs $10 less unless it's on sale. The Elite-X Fit 512GB is backed by a limited one-year warranty, which is a bit short considering that SanDisk offers a five-year warranty on the Ultra Fit.

NVMe drives offer the perfect balance of size and performance. They attach directly to a slot on most modern desktop and laptop motherboards and leave no mess of cables for us to hide away in our builds. 

PNY have just announced their latest range of M.2 2280 NVMe drives. Seen as an upgrade for users of SATA based SSD drives, the CS2130 range of drives provide an overall faster experience.

Coming in three capacities, 500GB, 1TB and 2TB. These drives offer up to six times the performance boost over older SATA SSD, thanks to the PCIe Gen 3 x4 interface.

But only the 2TB model offers the fastest sequential read speed of 3500MB/s and 3000MB/s sequential write. The 1TB model comes in at 3500MB/s sequential read and 1800MB/s sequential write. Which is no slouch for general desktop applications, but may hinder the performance of tasks requiring a large number of sequential writes. Sadly the 500GB model only offers up to 925MB/s sequential write speed, but still retains 3500MB/s sequential read. We were unable to determine if the drives use QLC or TLC as official information just lists it as "3D Flash Memory". An educated guess would be that these drives use QLC, given the consistent read speeds found across all of the drives. 

There are many other alternatives for high performance NVMe SSD drives but if you want a simple drive to rejuvenate an old desktop or laptop, then this may be an option. 

The drives start from $65 on Amazon for the 500GB model.

Graphics card vendor PNY has listed two new Quadro products on its website: the Quadro RTX 8000 Passive and Quadro RTX 6000 Passive.

Both graphics card flaunt a new design with the absence of a cooling fan for heat dissipation. Although the graphics cards sport a passive cooling system, it doesn't mean that they don't require active cooling. The Quadro RTX 8000 and RTX 6000 Passive are designed for data centers and workstations. They'll probably be installed inside specialized enclosures with high-speed chassis fans that can provide them with the necessary airflow, and therefore, can get away without having its own cooling fan.

The Quadro RTX 8000 and RTX 6000 Passive measure 10.5 x 4.4 inches and take up two PCI slots. One of the main advantages of a fanless design is that you don't ever have to worry taking your system apart to replace a dying fan.

The Quadro RTX 8000 and RTX 6000 Passive don't stray too far from their vanilla versions in regards to specifications. They retain the same TU102 silicon, featuring 4,608 CUDA cores, 576 Tensor cores and 72 RT cores. Nonetheless, the passive versions come with slightly lower core and memory clocks.

The Quadro RTX 8000 and RTX 6000 offer single-precision and Tensor performance up to 16.3 TFLOPs and 130.5 TFLOPs, respectively, while the passive variants are restricted to 14.9 TFLOPs and 119.4 TFLOPs, respectively. The memory bandwidth on the passive models top out at 624 GBps, while the non-passive ones check in with 672 GBps.

The main reason why the passive versions don't offer the same level of performance as the regular versions is because they are constraint to a tigher TBP (typical board power) envelope. The Quadro RTX 8000 and RTX 6000 are rated for 295W. The passive versions, on the other hand, max out at 250W. However, they still require a single 8-pin and 6-pin PCIe power connectors for external power.

As you would recall, the regular Quadro RTX 8000 and RTX 6000 are equipped with four DisplayPort 1.4 outputs and a USB-C port for Nvidia VirtualLink. Well, the passive variants don't come with any display outputs.

PNY backs the Quadro RTX 8000 and RTX 6000 Passive with a limited three-year warranty. The manufacturer didn't reveal the pricing for the graphics cards, but we expect it to be in the same range of the regular models.

Do you love taking photographs and recording videos on your smartphone, but always run out of storage? Are you sick of not having enough space for your Nintendo Switch games? Here's your opportunity to expand your storage space with PNY's U3 Pro Elite microSD card. The 256GB model, which retails for $139.99, is down to $59.99. The larger 512GB model that has a normal $189.99 is selling for $99.99.

• Get the PNY U3 Pro Elite 256GB microSD card for $59.99.
• Get the PNY U3 Pro Elite 512GB microSd card for $99.99.

Starting with the obvious, the PNY U3 Pro Elite is a microSDXC card that has earned the Class 10, U3 (UHS Speed Class 3), V30 (Video Speed Class 30) and A2 (Application Performance Class 2) certifications. In layman terms, PNY's offerings are fully capable of keeping up the latest 4K devices whether they be smartphones, action cameras, DSLRs and other devices.

Independent of the capacity, PNY's U3 Pro Elite microSD cards deliver up read and write speeds up to 100 MB/s and 80 MB/s, respectively. Whether you achieve the rated speeds depends on your device though. PNY also affirms that its U3 Pro Elite microSD cards are magnet-proof, shockproof, temperature-proof, waterproof and X-ray-proof. An SD adapter is included in case you want to use the U3 Pro Elite in a device with a standard SD slot.

They PNY U3 Pro Elite microSD cards are backed by PNY's limited lifetime warranty.

A sharp-eyed Redditor spotted a listing on PNY's website for its GeForce RTX 2080 Ti Overclocked XLR8 Edition for $1,000. The leak, which has since been taken down, also listed specs for the GeForce RTX 2080 8GB XLR8 Gaming Overclocked Edition.

The cards come bearing the Turing architecture. According to the listings, the $1,000 RTX 2080 Ti comes equipped with 4,352 CUDA cores and 11GB of GDDR6 with a 352-bit memory bus that provides 616 GB/s of throughput. PNY listed boost clocks at 1,545 MHz. The card is fed with two 8-pin connectors. However, it is also possible that the $1,000 price tag for the 2080 Ti is merely a placeholder. This same model also comes as a 1080 Ti that retails for $899, which potentially gives us an idea of the price deltas we can expect between generations.

The RTX 2080 comes packing 2944 CUDA cores and 8GB of GDDR6 with a 256-bit memory bus that pushes 448 GB/s of bandwidth. Boost speeds weigh in at 1,710 MHz. The card is fed with 8-pin and 6-pin power connectors. PNY purportedly listed this model for $800, but the listing was removed before we could grab a screenshot. Again, this is likely a placeholder.

As expected given Nvidia's recent announcements, the cards support AI and real-time ray tracing for games. Pictures of MSI's forthcoming models leaked earlier today, so it's clear the official launch is near.

PNY is ready to jump in the ultra-high-performance NVMe SSD market with the CS2080. The new M.2 2280 SSD comes with XLR8 branding, PNY's performance product group for gamers.

Performance is said to reach up to 3,200 MB/s sequential read and up to 3,000 MB/s sequential write speeds for the 1TB model. PNY will also release the CS2080 in 256GB, 512GB, and a massive 2TB drive that is sure to set your heart on fire. The series will ship with a three-year warranty.

The company didn't indicate when the CS2080 will ship, but the PS5012-E12 reference design shows strong performance in our early testing. We don't expect to see any drives with this controller until August even though some companies would like to reel that into the July time frame.

PNY plans to release the CS2080 add-in card form factor alongside the M.2 2280 models. The mock-up on display at Computex 2018 uses the same CS2080 part number. It had a simple sticker over the controller and Toshiba BiCS FLASH TLC. We hope the company will design a nice heatsink to give it a nice XLR8 look to match some of the company's other aggressive products.

PNY just announced the new 512GB Elite microSDXC memory card, which is large enough to take the capacity crown away from the 400GB device SanDisk released in February. For $350, mobile data hoarders can store over 80 hours of Full HD video or 100,000 18MP pictures.

The 512GB Elite satisfies the performance requirements to meet Class 10 and U1 ratings but not the hefty A1 requirement for application performance in a mobile device. PNY claims up to 90 MB/s performance over the UHS-1 interface but doesn't specify random performance, which is used to achieve the A1 rating.

Compatibility is always a concern for large capacity microSDXC products. PNY includes an SD adapter with the 512GB Elite for use in a PC or Mac but didn't publish a list of supported mobile devices.

PNY is the first to announce a 512GB microSD card, but we know more will come in a few weeks at the Computex trade show in Taipei, Taiwan. The increase in density is a direct result of advances in flash technology. PNY didn't disclose the technology used to make the 512GB Elite microSDXC possible but it's likely the result of new 4-bit per cell memory (QLC) from Toshiba.

What worries us about the 512GB Elite microSDXC is the lack of any published endurance specifications. QLC is new, unproven memory technology for consumer devices and will require powerful software error correction technology that increases power consumption over the life of the device.

Introduction, Technical Data & Details

PNY's PrevailPro mobile workstation family starts in the $2500 range, but quickly ramps up to the around-$4500 P4000 model we're reviewing today. Availability seems to be working itself out slowly. Though the PrevailPro family was announced in November of 2017, a handful of online vendors do show configurations ready to ship within a couple of days.

A closer look at the tooling indicates that PNY is using the Taiwanese ODM Clevo as a contract manufacturer. PNY's version is based on the P955 platform, but it enables more opulent hardware options. After all, the company needs to justify its business-oriented price tag.

The PrevailPro Base and Upgraded Pro both sport lower-end graphics and storage options, saving hundreds of dollars in the process. But as today's review will show, the flagship's Core i7-7700HQ and Quadro P4000 mobile GPU are a well-balanced pair. Obviously, PNY's motto here is go big or go home.

The Bundle

In addition to the notebook, you get an external 150W power supply, a quick-reference guide, a microfiber cloth for removing fingerprints, a DVD, and a USB stick. Since the PrevailPro doesn't even have a DVD drive, the disc probably could have been left out.

Look & Feel

A width of 14.96 inches, a 9.8-inch depth measurement, and a closed z-height of 0.73 inches make this a fairly average-sized device, given its 15.6" display. The PrevailPro does weigh 4.8 pounds, putting it on the heavier side. But obviously a high-end workstation's performance (and cooling) has to come from somewhere. We're just glad it's not an all-plastic affair bound to break sooner than later.

Unfortunately, the metallic matte surface covering everything from the lid to the base plate, as well as the interior plastic surfaces, is extremely sensitive to oily fingerprints.

Connectivity & Audio

With a few exceptions, the PrevailPro is very similar to Clevo's base platform. The left side hosts a security lock slot, ventilation for the CPU cooler, a DC power-supply connector, one HDCP-compatible HDMI 2.0 output, two mini-DP 1.3 interfaces, two Type-C USB 3.1 (Gen 2) ports, and two USB 3.1 (Gen 1) ports. It really wouldn't be possible to cram more connectivity on that side.

On the right, you get a 3.5mm headphone jack, a combined 3.5mm 2-in-1 audio port (microphone and S/PDIF), one USB 3.1 (Gen 1) port, a 6-in-1 card reader, and an RJ-45 GbE port.

The back is dominated by three cut-outs for ventilation. Two are for the graphics processor, and one allows general passive exhaust.

The front is mostly unpopulated, except for the left side with its status LEDs.

Up close, those LEDs include the operating status and mains indicator, battery status, storage activity, and airplane mode.

Beyond the wired GbE controller, connectivity also includes Intel's dual-band Wireless-AC 8265 (M.2 2230) module. Right next to a Fritz!Box 7590 router, we measured a maximum of 674 Mb/s. Range and data rates in the laboratory correspond to those of a proper USB-attached Wi-Fi dongle, even though internal antennas do have some disadvantages in terms of range. We didn't notice any real disruptions or connection breakdowns.

Furthermore, the Windstorm Peak module incorporates Bluetooth 4.2 support as part of a well-rounded communications package. It does nibble away at battery life if you don't switch it off, though.

The two built-in loudspeakers are sufficient for a workstation, but we'd stop short of calling them impressive. Integrated audio, enabled through a Creative Sound Blaster X-Fi BM5, supports high-definition audio, as well as ANSP 3D sound at the output. With headphones attached, the sound is quite nice. To be sure, this platform has its roots in the gaming segment; the motherboard employs nearly identical components across all derivative implementations.

Input & Output

The keyboard utilizes standard-fare short-stroke keys, and its configurable backlighting is quite nice. However, this is probably more of an inherited feature of Clevo's platform rather than something we'd choose to see on a mobile workstation.

Large (but still soft) rubber pieces on the enclosure's bottom side ensure a stable grip on the surface below.

Our sample employs a matte Ultra HD IPS panel made by LG Philips (LP156UD1-SPB1), which features a 60Hz refresh rate, but isn't particularly state-of-the-art anymore. In contrast to pure RGB panels, this RGBW version makes use of a sub-pixel array that, under certain circumstances, may have a negative impact on contrast. Black areas in particular are just not as black as they could be. Although backlight bleeding does occur, we noticed only mild effects. The screen's four corners are affected, as are spots on the upper and lower edge. Depending on the display's angle and on-screen content, this is noticeable but still tolerable. Maximum brightness attached to mains power is sufficiently high: ~260 to 300 cd/m².

Upcoming retail models will come with an AUO B156ZAN02.2. These are AHVA RGB panels, like those already used by other Clevo-made notebooks. A 92% coverage measurement of the sRGB color space is fine, but the meager 62% for AdobeRGB is merely satisfactory for a workstation in this price range.

The display can be tilted up to 130 degrees. This is more than enough, and it doesn't make the device top-heavy or unsteady, either.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Tear-Down, Hardware & Cooling

In this case, a tear-down is necessary because PNY doesn't provide detailed specs on all of the PrevailPro's hardware. For example, we didn't know what memory the company used, nor did we have a make/model on the SSD.

To open the notebook and get a look inside, it's first necessary to loosen two screws in back holding the keyboard in place. Next, nudge the keyboard away from the chassis by pushing it from behind through one of the screw holes. This allows the keyboard to be carefully removed from its mount. Once this is done, the two connecting cables can be pulled out using the designated lugs.

The tactile feedback of the keys is noticeable in both feel and sound. However, they're still relatively soft. Typing on the keys is comfortable thanks to large surfaces and plenty of space between them. A little more resistance would make the keys even better, we think. Unfortunately, this form factor's height limitations prevent the use of more-tactile scissor keys.

Included FlexiKey software provides customizable macros, keyboard recording options, and a three-zone RGB backlight. The absence of LED indicators for caps and number lock is a bit unfortunate.

Next, we loosen five exposed screws underneath the keyboard that hold the top to the bottom, close the notebook, and carefully turn it on its back to remove the remaining nine screws. After that, the bottom part of the case is reasonably easy to remove, exposing the platform's internals.

This is a fairly classic configuration, with the GPU and CPU cooled separately. The battery pack and other modular components are now accessible for easy replacement.

Component Analysis

For memory, PNY uses OEM RAM modules made by Samsung. Two 16GB DDR4-2400 SO-DIMMs add up to 32GB in dual-channel mode. Together, they max out the platform's capabilities; you cannot add any more than 32GB of of memory.

PNY doesn't provide much information about the SSD, either. Our tear-down reveals that it's actually a Samsung PM961 featuring the following specs:

This is a desirable model that was originally designed for OEMs only, but it can still be purchased online. As far as pricing goes, however, there aren't really any advantages versus Samsung's standard consumer models.

PNY's Wi-Fi module of choice is the Intel Dual-Band Wireless-AC 8265. Two connected cables for the antennas are visible on the right side.

Cooling

Components from Auras Technology, one of the largest heat-pipe suppliers, are used for the thermal solution. The blackened copper heat sink and composite pipes making up the GPU cooler are arranged as a single module. Air flow and exhaust are provided by a pair of radial fans that suck air in from the enclosure's bottom and push it through two finned chamber coolers. Warm air then leaves the chassis through an opening on the back.

The CPU is cooled separately. Its module is made by Auras, as well, though two heat pipes and a shorter cooling chamber must suffice for this lower-power component. Heated exhaust is blown out the notebook's left side, not the rear. Since most users are right-handed, Clevo's design ensures they aren't (unpleasantly) affected by the hot air.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Benchmarks: 2D & CPU-Heavy Workloads

System Setup

To what do you compare a mobile workstation when it's just a normal consumer notebook with Quadro graphics? The Intel Core i7-7700HQ does not support ECC memory, and even though the other components are workstation-like, PNY's PrevailPro P4000 is built almost exclusively with standard hardware.

In order to keep the CPU at least somewhat similar, we bought an Intel Core i7-7700T with a slightly higher 2.9GHz base clock rate and an equal 3.8GHz Turbo Boost ceiling. We installed 32GB of Kingston DDR4-2400 and dropped it onto an MSI Z270 Gaming M7 with a M.2-attached Samsung 960 Pro 512GB SSD.

For graphics, we tapped the Quadro P4000 and P2000 to benchmark against. The mobile Quadro P4000 has a slightly lower GPU frequency and a 400MHz memory deficit. Meanwhile, the Quadro P2000 employs a GP106 processor instead of GP104, giving us a more complete performance picture. According to its technical specifications, the mobile Quadro P4000 Max-Q should end up somewhere in the middle, perhaps with a tendency toward the Quadro P4000.

Benchmarks

You'll notice that we're avoiding the usual synthetics, since targeted driver optimizations often tell an unrepresentative story. Instead, we rely on licensed and certified workstation software. After all, enthusiasts who wish to render movies or run DirectX-based suites can find better (and less expensive) options in the consumer space.

CPU-Heavy 2D Graphics Output

Using our GDI/GDI+ benchmark, we first test two different output methods for 2D objects that are typical of older applications. They are still used today either to generate user interfaces or for printing purposes.

To this end, the graphics driver uses the CPU quite extensively, but typically occupies only a few threads. After all, there hasn't been any real 2D hardware acceleration since the Unified Shader Model was introduced. The Windows driver model really puts the brakes on this, as well. Since every single action has to pass through the entire system, we also get a good depiction of platform performance.

Next, we introduce memory to the game as we use the only remaining 2D hardware function: copying graphics created in memory to an output device. This follows the same procedure, except all drawing is done on a virtual bitmap first, rather than going straight to the monitor. Then, only the complete graphic is sent to the output device via bit block transfer. This time around, the graphics card plays a decisive role as the CPU takes a back seat.

The 2D AutoCAD 2017 test exposes shortcomings in IPC throughput, making this a predominantly theoretical measurement and the logical continuation of our two previous synthetic benchmarks.

The CPU composite score in SolidWorks shows very clearly that both single- and multi-threaded scenarios influence the overall result to some extent, although rendering performance contributes to the outcome, as well. Thus, the graphics card also plays a role, even if the actual focus is on host-processing performance.

The same applies to PTC's Creo 3.0. After all, it is, like SolidWorks, another important piece of software in the development field.

The CPU composite for 3ds Max is significantly less dependent on graphics performance, revealing the Core i7-7700T-based system's advantage. It is only Intel's feeble HD Graphics engine that drags down the results.

Next up, we have a CPU-intensive workload that couldn't be more different. It calculates the yields and losses of a complex photovoltaic system based on long-term weather data and sun positions, including possible errors and power drops during operation. A total of eight threads creates an almost 100% load. In this case, 3D graphics output to visualize intermediate results has only marginal impact on the overall result. PNY's PrevailPro is definitely at a disadvantage.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Benchmarks: Complex Workloads

Now let's combine CPU-intensive and real-time graphics tasks. The nominally faster Quadro P4000 Max-Q is at a disadvantage against the desktop Quadro P2000, which benefits more fully from GPU Boost.

We start with SolidWorks 2015's GPU composite, along with three selected individual results (sub-composites).

A similar picture is painted by PTC's Creo 3.0. The PrevailPro's lower CPU performance limits its theoretically faster GPU.

Even in Adobe's Creative Cloud, the CPU's influence is unfortunately what prevents the Quadro P4000 Max-Q from performing better.

It may look like the following benchmark is graphics-limited. However, we also registered high CPU utilization in one thread. Not surprisingly, then, the notebook once again suffers a disadvantage.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Benchmarks: Graphics-Heavy 3D (OpenGL)

To keep everything balanced, we're also interested in looking at workloads where the Quadro P4000 Max-Q is allowed to operate almost entirely unrestrained. In those types of scenarios, a Quadro P2000 add-in card often trails.

When we load a complex scene in Maya 2017 and simply rotate it around, the mobile Quadro shines, landing precisely between the two add-in workstation cards.

Since its CPU load is even lower than Maya 2017's, 3ds Max shows us that the mobile Quadro P4000 Max-Q can land closer to Quadro P4000 than the P2000.

AutoCAD's DirectX functions are no challenge for modern GPUs. Maybe that's why Intel's HD Graphics engine fares better here than in previous benchmarks.

One 80%-utilized CPU thread has a noticeable impact on performance, causing the mobile Quadro P4000 to land down by the add-in P2000.

Results from Cinebench R15's OpenGL benchmark are even more extreme, though we're able to blame drivers for once. Remember when we mentioned targeted optimizations favoring benchmarks? This workload means very little when it cannot be extrapolated to other tasks.

The mobile Quadro P4000 Max-Q is adequate, but limited by the Core i7-7700HQ in applications that would have otherwise been GPU-bound. Whenever the CPU comes under heavy load, a desktop Core i7-7700T is able to propel the nominally weaker Quadro P2000 beyond the theoretically faster mobile solution.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Power Consumption, Battery Life & Noise

Power Consumption

Power use varies depending on whether you're plugged into the wall or using the battery. But since we can't modify this sample, we're only able to measure between the mains socket and power supply. Thus, the following table is limited to consumption on mains power.

Our tests include several scenarios, including sleep mode. The stress test is rather theoretical in nature; you'll never replicate it using real-world applications.

Real-life measurements, even for something like a larger SolidWorks workload, make it obvious just how important power-saving technologies are to stretching out battery life. Still, don't expect to run demanding applications away from the wall for any length of time.

Even though CPU and graphics performance is dramatically reduced on battery power, taking consumption down a notch, there's only so much capacity to keep this platform running. We keep its display on the normal brightness level across all workloads except the stress tests, where we max it out.

Once we completely deplete the battery, it takes less than two hours to fully recharge it, assuming you don't use the notebook during that time.

Noise

Clevo employs three fans in this design, similar to its original P955 platform. The graphics module's thermal solution consists of three flattened heat pipes connected to a sink with two radial fans, while the considerably more economical CPU gets by with two heat pipes and one radial fan. Noise generated by both subsystems under load is quite audible.

We even pick up a low humming sound when the PrevailPro is idle and charging. Higher-quality coils for the DC-DC converters would have helped there.

From 50cm away, our equipment registers approximately 40 dB(A) through a battery-powered SolidWorks benchmark run that pushes 2D/3D, compute, and rendering. This is an acceptable result.

But the almost 46 dB(A) we measured from the same run at higher clocks (plugged into the wall) is a bit too much. The experience is all the more uncomfortable as the keyboard above and your legs below start heating up.

Under maximum load, the notebook gets up to almost 50 dB(A), which is easily enough to drown out any other equipment in your office. The three fans spinning as fast as they can definitely take a toll. After all, up to 150W of energy converted into heat in a compact enclosure has to be dealt with one way or another.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Temperatures, Clock Rates & Thermals

Temperatures & Clock Rates

Actual performance typically differs from manufacturer specifications, if only because frequencies dip once a platform warms up. Because the PrevailPro is a workstation-class notebook, it's reasonable to assume that professionals using it for work will be plugged into a wall outlet for their heavy lifting. As such, that's how we'll test.

During a demanding SolidWorks run, the average CPU clock rate stays just below 3.5GHz with a load well-distributed between cores. We don't actually reach the peak Turbo Boost frequency at any point. CPU core temperatures range between approximately 60°C and 70°C. The Quadro P4000 Max-Q maintains an average of 1100MHz to 1250MHz across a temperature range of about 50°C to 56°C, depending on the somewhat different workloads in our benchmark.

During our stress test, the CPU drops below 3.4GHz and rises above 85°C, but is able to maintain those values. The GPU eventually stabilizes around 69°C and achieves a frequency of about 1100MHz.

There is a somewhat larger gap between the maximum Turbo Boost and GPU Boost frequencies, and what we actually observe. But this delta is something we'd consider to be generally inherent notebook behavior, too. Naturally, this makes direct comparisons to desktop PCs with well-cooled components difficult.

Temperature Measurements

Accurate measurements are a prerequisite for any fair evaluation. In order to keep our results as representative as possible, we keep the chassis intact (rather than zeroing in on specific components inside). Furthermore, we covered the opening in our workbench with special foil that we tested beforehand to determine its thermal permeability. Thus, we're able to ensure that our approximation of a normal user environment is as realistic as possible. (An earlier test with exposed platform components resulted in significantly lower numbers, so we discarded that data.)

We also noticed that it does not help to simply rest the opened notebook in an upright position against a surface. On the contrary, this causes the heat pipes to either lose performance or stop working altogether. Consider that a word of warning if you were thinking about using your notebook on a stand at too steep of an angle. Up to about 25° is safe, but we wouldn't suggest going any further.

After finding the correct focus point, we took measurements right through ventilation openings on the back, which gave us a good view of the GPU cooler's lowest (and hottest) heat pipe. We also have readings from the laptop chassis' surface.

During normal operation and through our SolidWorks benchmark, readings remain in a tolerable range. We also measure an average delta of seven to eight Kelvin between the heat pipe and GPU diode, which is perfectly fine. The CPU runs a bit warmer, but that has less of an impact on the overall thermal load.

Even though a stress test on battery power isn't particularly practical, it does yield interesting results. The notebook operates at roughly the same level as it did in SolidWorks, though MSI Kombustor coaxes the GPU into generating a little more waste heat, despite obvious throttling to keep thermals in check.

Of course, the battery won't last even an hour if you try to use it like this.

In order to determine a worst-case scenario, we plug in a power cord and crank up the display's brightness. A reading around 65°C won't cause anything to melt, but it's certainly enough to get your legs nice and toasty, as the laptop's bottom panel gets up over 44°C.

We also want readings from the top side, since you're always going to have your hands on the keyboard during use. Rather than using measuring lacquer for our thermal imager or thermal adhesive for sensors, we went with an infrared thermometer to capture data from the top side.

Connected to a wall outlet, our stress test causes temperatures in the center-upper quarter of the keyboard to rise to 54°C. The F8, 8, and I keys get the warmest. In a 3D workload like SolidWorks, temperatures are still around 46°C to 48°C. Otherwise, the key caps simply stay a little warm.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Conclusion

PNY does almost everything right with its PrevailPro P4000. The mobile Quadro P4000 Max-Q design works well with Intel's Core i7-7700HQ CPU. Although Kaby Lake as an architecture is no longer state-of-the-art, it's still modern for mobile platforms. Besides, Clevo relies on a mature P955/957 platform to make the PrevailPro P4000's unique component list possible.

Aside from the Quadro graphics module and big DDR4 SO-DIMMs, everything else inside is consumer-grade hardware. Then again, ECC RAM could be considered unnecessary in a mobile platform, and an expensive mobile Xeon CPU to drive the Quadro wouldn't help much. The combination in PNY's PrevailPro already makes a nice team.

Whether an Ultra HD panel makes sense while traveling is another matter entirely. Many professional applications, such as AutoCAD, Adobe's CC, SolidWorks, and Creo rely on their own GUIs that unfortunately scale with the screen. This makes them very difficult to read and use. Attempts to install SolidWorks 2015, for example, were certainly not enjoyable. Even scaling font sizes up to 300% is no use when dealing with a GUI that relies on fixed-size bitmap elements.

While PNY's PrevailPro P4000 is certainly well-balanced, we're left wondering about the application that calls for such a notebook. A traveling engineer working on construction plans and presentations would be on a perpetual hunt for the next power outlet, and as a primary device docked wherever you need to work, the acoustics aren't particularly optimized. You'd need a strong voice and plenty of patience to work with all of that noise.

On the other hand, we have to admire the cooling subsystem's effectiveness. This platform can handle the most powerful graphics solutions in a compact form factor, even if it means blowing lots of air over them. CPU cooling isn't a concern at all in comparison. Although you won't see the peak Turbo/GPU Boost frequencies in most professional workloads, throttling isn't an issue for PNY's PrevailPro P4000. Performance is maintained, whether you're plugged into the wall or running off of battery power.

We don't see enough unique selling points to issue PNY's laptop an award. However, the notebook does exactly what it promises if you happen to need its Quadro graphics module on the road. That alone makes this a unique piece of hardware if you're in the target market for its professional GPU. A $4500 asking price is higher than we'd be willing to pay. But if someone in accounting gives you the green light for a big expense, there are legit reasons to favor it over gaming-oriented notebooks.

MORE: Best Gaming Laptops

MORE: Gaming Laptop Previews

MORE: All Laptop Content

Just a few short days ahead of Computex 2017, PNY has taken the wraps off of its GeForce GTX 1080 Ti XLR8 Gaming OC graphics card. (Get it? XLR8? Accelerate? You get it.)

In addition to sporting all the features you've come to expect from GeForce GTX 1080 Ti-based cards, such as 3,584 CUDA cores, 11GB GDDR5X, 352-bit bus, 224 texture units, 28 streaming multiprocessors, 88 ROPs, and a 250 TDP, PNY stated that its upcoming flagship graphics card is put through a "rigorous" binning process to make sure each GPU is able to achieve the advertised 1,645MHz boost clock speed.

The cherry on top of the XLR8 cake is the massive triple fan cooler. Under the large red and black shroud, you’ll find three fans pushing air through interlaced aluminum fins bonded to four 6mm nickel-plated copper heatpipes that are attached to a large copper baseplate. The anodized aluminum backplate spans the entire length of the graphics card, adding rigidity and structural strength to prevent it from bending under the weight of the beefy XLR8 cooler.

Surprisingly, the XLR8 triple fan cooler adds just $15 to the price of this graphics card compared to the company's GTX 1080 Ti GPUs that are outfitted with standard blower style coolers.  

The PNY GeForce GTX 1080 Ti XLR8 Gaming OC is available now for $725 on the company website.

Introduction & Overview

Since we published our GeForce GTX 1080 review, Nvidia overcame availability issues, and prices on the top-end gaming card settled into the range we were told to expect at launch. Today, you'll find GeForce GTX 1080 cards selling online between $600 (£750) and $770 (£900).

When Nvidia introduced GeForce GTX 1080, we only had the Founders Edition board (the company's reference design) in our possession. While its rear exhaust and a high-quality thermal solution turned heads, we knew boards from Nvidia's partners could bring lower prices, factory overclocking, and more cooling designs to the table.

All of these cards employ the same GP104 processor, so gaming performance takes a back seat to acoustic, electrical, and thermal readings. Each board is reviewed on its own page, where we dissect build quality, differentiating features, power consumption, clock rate analysis, cooling, and acoustic measurements in depth.

Including Nvidia's GeForce GTX 1080 Founders Edition, we have 10 cards compared in this roundup so far. We'll continue adding GeForce GTX 1080 cards to this roundup as we review them.

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

The final analysis of each card is listed below for quick and easy comparison. The cards appear in alphabetical order, and do not reflect any valuation or preference for individual products.

MORE: Best Deals

Benchmark Results

Gaming

We test every card after a suitable warm-up period to avoid unfair differences in GPU Boost frequencies. All benchmarks are run six times; the first one is used to get the GPU hot again.

These cards are all press samples operating at the same settings as retail models in our best effort to ensure one vendor doesn't get a leg up on another using non-representative clock rates.

The following galleries each contain four images, covering two tested resolutions. We put our focus on QHD (2560x1440) and UHD (3840x2160), plotting out average and minimum frame rates for each resolution in separate graphs.

All of the factory-overclocked cards offer similar performance, more or less. That's why our primary focus centers on evaluating the more technical aspects of each board design, along with their coolers. This is where differences in noise, power, and temperatures are most likely to come from.

Power Consumption

We begin by comparing the power consumption of each card in our gaming loop, stress test, and at idle. Depending on the power targets specified by each manufacturer, we sometimes measured substantial differences. We're ignoring the decimal places in our bar graphs, since those values are too small and would be within measurement tolerances.

We also found that some cards with lower power targets started to throttle during our stress test, resulting in lower power consumption numbers. This did not, however, have a negative impact on general gaming performance for any of the tested cards, as the stress test merely represents a worst-case scenario.

We confirmed that MSI's retail cards will ship with a slightly lower power target (max. 240 to 250 watts) after an internal discussion and evaluation of our measurements.

This also applies to the BIOS versions with OC mode enabled by default, which employ a roughly 20 MHz-higher base and GPU Boost frequency. In the interest of fairness, we tested both MSI cards using normal mode, without the overclocked base and GPU Boost rates. This doesn't affect our power consumption measurements, though.

Noise

For the following comparison, we divide all of the gallery's bar graphs into gaming loop and idle, even if "noise" in practice spans a wide spectrum and the "character" of the sound varies a lot, with each card having highly individual results. Therefore, it is important not to compare just the absolute numbers, but also the frequency spectrum we're presenting.

Many of the cards implement a semi-passive mode, where their fans remain off when the card is idle. Thus, we refrained from taking measurements in that state. Even in our anechoic chamber, levels of 22 dB(A) and below merely represent ambient noise.

Temperature

For this comparison, we divide the gallery's bar graphs into gaming loop, stress test, and peak temperatures measured on the MOSFETs.

We occasionally compared the temperatures on our benchmark table with those measured inside a closed case and found them to be no more than two or three Kelvin higher.

Since temperatures in a closed case also depend heavily on the enclosure's cooling performance, the only representative and reproducible values are those measured on our benchmark table. Those are the ones we compare.

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Nvidia GeForce GTX 1080 Founders Edition

The GeForce GTX 1080 Founders Edition is Nvidia's current desktop flagship by its own design. Despite an emphasis on craftsmanship in the company's marketing materials and generally improved efficiency, the reference design will have a hard time trying to justify its high price compared to compelling solutions from add-in board partners. This is doubly applicable since the restrictive power target of just 180W and a 172°F (80°C) temperature target are real limitations.

As with the GeForce GTX 1070 Founders Edition, Nvidia chose a mechanical-looking design and a radial fan. The card's true two-slot form factor also makes it a good option in multi-GPU machines.

Technical Specifications

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Exterior & Interfaces

The GeForce GTX 1080 Founders Edition's shroud is made of injection molded aluminum, and colored both silver and black. The metal cover coveys plenty of quality, but also results in a weight of more than 35 ounces (one kilogram). Both this 1080 and Nvidia's GeForce GTX 1070 Founders Edition have an almost identical weight (the 1080 is just one-third of an ounce, or 10 grams heavier).

Around back, the board is covered by a two-piece plate that serves no practical purpose other than to facilitate a more finished appearance. If you want to increase airflow in a multi-card configuration with 1080s back to back, the plate can be unscrewed without causing a problem.

The top of the card is dominated by a glowing, green GeForce GTX logo. The eight-pin auxiliary power connector is positioned at the end of the card. The sharp-edged and mechanical design may be a matter of taste, but it certainly is distinctive.

A peek at the card's back reveals fins and a mounting frame. Three screw holes are provided for attaching brackets to stabilize the card in a case.

Five outputs populate the rear bracket, four of which can be used simultaneously in a multi-monitor setup. You get one dual-link DVI-D connector, one HDMI 2.0b port, and three DisplayPort 1.4 outputs. The rest of the plate is ventilated to maximize exhaust flow.

Board & Components

A glance at the PCB reveals that it offers significantly more space than is actually used. In addition to one power phase for the memory, five of the six possible phases for the GPU are implemented. There's even space for an extra power connector, if it's needed. 

As you no doubt know, Nvidia taps GDDR5X memory from Micron for this board. Eight of these memory chips operating at 1251 MHz are connected to the GPU through an aggregate 256-bit interface, enabling a theoretical bandwidth of 320 GB/s.

The 5+1-phase system relies on the sparsely documented µP9511P PWM controller. Since this controller can't communicate directly with the VRM's phases, Nvidia utilizes 53603A chips for solid PWM drivers (gate drivers) and controlling the power MOSFETs (primarily of type 4C85N).

Two capacitors are installed right below the GPU to absorb and equalize peaks in voltage. The board design looks tidy and well thought-out.

Power Results

Before we look at power consumption, we should talk about the correlation between GPU Boost frequency and core voltage, which are so similar that we decided to put their graphs one on top of the other. This also shows that both curves drop as the GPU's temperature rises.

After warm-up and under load, the GPU Boost clock rate at times drops to GP104's base frequency of 1.605 GHz. This is mirrored in our voltage measurements. While we measured up to 1.062V at first, that number temporarily drops as low as 0.881V.

Combining the measured voltages and currents allows us to derive a total power consumption we can easily confirm with our instrumentation by taking readings at the card's power connectors. In fact, let's start with the measured power consumption values in the following table:

These charts go into more detail on power consumption at idle, during 4K gaming, and under the effects of our stress test. The graphs show how load is distributed between each voltage and supply rail, providing a bird's eye view of load variations and peaks.

Temperature Results

To dissipate the GP104 GPU's waste heat, Nvidia's GeForce GTX 1080 Founders Edition employs a real vapor chamber solution. The 1070 Founders Edition does not benefit from this same technology. The vapor chamber is a compact module that's attached to the PCB via four screws and positioned over the GPU package.

A radial fan pulls air from inside the chassis and blows it across heat sink fins on the vapor chamber, exhausting that air through the output bracket. The mounting frame is not only used to stabilize the card, but it also helps cool the voltage regulators and memory ICs.

The vapor chamber's performance is slightly better than the 1070 Founders Edition's copper heat sink, as you might expect. Still, temperatures rise to almost 185°F (85°C) during our stress test and 180°F (82°C) during the gaming loop. Then again, that's hardly shocking since the 1080 is rated 30W higher.

Sound Results

Initially, both fan speed curves stay relatively close to one another. But at a certain point in our benchmark, FurMark forces the RPM measurement to burst much higher. During the more real-world gaming loop, however, the fan cruises along just above 2100 RPM.

Noise levels are relatively low when the card is idle, even if the fan's sound has a slightly snarly character. Nvidia doesn't gift the 1080 Founders Edition with a semi-passive mode. But then again, it wouldn't be particularly useful on a card with a radial fan anyways.

Even under longer gaming loads, noise levels stay below the 42 dB(A) mark. Not bad for a radial fan. However, during our purposely taxing stress test, the fans ramp up beyond 46 dB(A). The frequency spectrum is rather wide though, so the white noise doesn't feel too intrusive.

Overall, Nvidia's thermal solution is workable. The radial fan is great for exhausting hot air from the card's back, but it's miles away from making the 1080 Founders Edition quiet.

MORE: Best Deals

MORE: Hot Bargains @PurchDeals

EVGA GeForce GTX 1080 FTW Gaming ACX 3.0

FTW is the shortened version of For The Win, and with that abbreviation EVGA sets the bar pretty high for itself. We do like that the card can be disassembled without voiding its warranty. The terms of EVGA's coverage are thus very favorable to water-cooling and modding enthusiasts. This is downright uncommon, which is why we want to point it out upfront. But we still don't know anything about the 1080 FTW Gaming ACX 3.0's technical attributes...yet, at least.

Technical Specifications

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Exterior & Interfaces

The cooler cover is made of light metal. Together with the underlying acrylic plate, the extra tacked on are primarily designed to provide some eye candy. As a result, the whole top of the card looks like it's bathed in color; built-in RGB LEDs light up the shroud thanks to numerous recesses. Right out of the gate, EVGA's FTW is the brightest card in our test field.

Weighing 38oz (1077g), this card isn't particularly heavy. But it's not a lightweight either. Measuring 11 inches (or 27.7cm), it isn't excessively long. It's five inches (or 12.5cm) tall, though, and 1 3/8 inch (3.5cm) wide, matching many dual-slot graphics cards.

The back of the board is covered by a single-piece plate with several openings for ventilation. You'll have to plan for an additional one-fifth of an inch (5mm) beyond the backplate, which could negatively affect configurations with multiple cards right next to each other. It is perfectly possible to use the card without that cover, of course. However, removing it also necessitates pulling off the cooler.

The top of the card features EVGA's logo. Two eight-pin power connectors are positioned at the end of the card, right where we'd expect to find them. As with most designs that strive to be unique, this one is a matter of personal preference. We're sure it'll find its fans, though. While there are undoubtedly fancier cards available, being fancy isn't always a compliment either.

Fins visible from the end and bottom of the 1080 FTW Gaming ACX 3.0 show that they're positioned vertically, and won't allow air to flow out of the back. Instead, exhaust is pushed from the top and down, toward the motherboard.

The rear bracket exposes five connectors, four of which can be used simultaneously in a multi-monitor setup. In addition to one dual link DVI-D connector, you also get one HDMI 2.0b port and three full-sized DisplayPort 1.4 outputs. Ventilation holes dot the rest of the bracket. They don't serve any purpose though, given EVGA's fin orientation.

Board & Components

EVGA leverages its own circuit board design. At first glance, the card seems to employ a very clean layout that sparks our curiosity due to its many on-board components.

EVGA naturally uses the Micron GDDR5X memory modules that Nvidia sells along with its GPU. Eight of them operating at 1251 MHz are connected to GP104 through an aggregate 256-bit interface, enabling up to 320 GB/s of bandwidth.

Unlike Nvidia's reference design, EVGA's 5+2-phase system relies on an NPC81274 from ON Semiconductor as its PWM controller. Even if the card is advertised as having 10 power phases, that's technically a little deceiving. In reality, there are five phases, each of which is split into two separate converter circuits. This isn't a new trick by any means. It does help improve the distribution of current to create a larger cooling area. Furthermore, the shunt connection reduces the circuit's internal resistance. This is achieved with a NCP81162 current balancing phase doubler, which also contains the gate and power drivers.

For voltage regulation, one highly-integrated NCP81382 is used per converter circuit, which combines the high-side and low-side MOSFETs, as well as the Schottky diode, in a single chip. Thanks to the doubling of converter circuits, the coils are significantly smaller. This can be quite an advantage since the current per circuit is smaller as well. As a result, conductors can be reduced in diameter while retaining the same inductance.

Yet, compared to the KFA²/Galax GeForce GTX 1080 HoF, which uses this effect for a total of 15 converter circuits, EVGA achieves rather modest results as far as cooling is concerned. As we get into the 1080 FTW's benchmark results, we'll present some data backing that claim.

Current monitoring is enabled through a three-channel INA3221.

The memory modules are powered by two separate phases controlled with a NPC81278, which integrates the gate driver and PWM VID interface. A NTMFD4C85N by ON Semiconductor combines both high-side and low-side MOSFETs in one chip.

Two familiar capacitors are installed right below the GPU to absorb and equalize voltage peaks.

Power Results

Before we look at power consumption, we should talk about the correlation between GPU Boost frequency and core voltage, which are so similar that we decided to put their graphs one on top of the other. EVGA uses a very high power target that, in turn, allows for a relatively constant GPU Boost clock rate. It only drops slightly as temperature increases, and the observed voltage behaves similarly.

After warming up in our gaming workload, the GPU Boost frequency, which initially started at 2 GHz, settles at a stable value of 1936 MHz. This falls to 1848 MHz under constant load.

Our voltage measurements look similar: while we observe up to 1.062V in the beginning, that number dips to an average 1.031V.

Combining the measured voltages and currents allows us to derive a total power consumption we can easily confirm with our instrumentation by taking readings at the card's power connectors.

Since Nvidia forces its partners to sacrifice the lowest possible clock rate in order to gain an extra GPU Boost bin, this card's power consumption is disproportionately high as it idles at 253 MHz. EVGA handles this setback fairly well, though. The consequences are listed below:

These charts go into more detail on power consumption at idle, during 4K gaming, and under the effects of our stress test. The graphs show how load is distributed between each voltage and supply rail, providing a bird's eye view of load variations and peaks.

Temperature Results

Update, 11/26/16: In the original version of this round-up, we exposed an issue with EVGA's ACX 3.0 thermal solution. The company used our feedback to improve its design, in the process giving customers a couple of options for modifying their existing cards. What follows is a review of those options and their impact on cooling/noise. We're replacing the previous data with results from our updated testing, since that most accurately reflects production hardware.

Naturally, heat output is directly related to power consumption, and the 1080 FTW's ability to dissipate that thermal energy can only be understood by looking at its cooling solution.

As with the 1080 Founders Edition, the backplate is mostly aesthetic; it doesn't serve much practical purpose. At best, it helps with the card's structural stability.

Unfortunately, the way the voltage regulation circuitry was originally cooled exposed a design flaw. This problem was evident in the mounting plate, which does draw heat away from the VRMs and memory. The plate has a cut-out for the coils, meant to keep the VRMs and memory from heating each other up. Unfortunately, it also reduced the cooling surface for an area right below the center of the fan rotor, where airflow is weak to begin with.

Under normal circumstances, the company's decision might have still turned out alright. But the GeForce GTX 1080 FTW Gaming ACX 3.0 has an exorbitantly high power target, which chases the card up to and beyond 230W under load. This puts a lot of stress on the VRMs. Given those conditions, the limited surface area just couldn't keep up.

The cooler uses a nickel-plated heat sink, along with four quarter-inch (6mm) and two one-third-inch (8mm) flattened heat pipes. Its actual capacity is adequate. But a look at the GPU temperatures shows that the upper limit of this short, dual-slot solution is in sight. We originally measured up to 167°F (75°C) during our gaming loop and up to 172°F (78°C) in a closed case. Under maximum load, the reading rose to 171°F (77°C), and 176 to 178°F (80 to 81°C) in a closed case. That was just too high, especially when you consider the data was collected in an air-conditioned room set to 72°F (22°C).

Re-Testing After EVGA's Thermal Modification

Clearly, something needed to be done. Back in November, we published EVGA Addresses GeForce GTX 1080 FTW PWM Temperature Problems, linking to new BIOS versions for five different cards. EVGA also offers optional thermal pads to any affected customer who requests them. Customers who do not feel comfortable updating the BIOS or who damage their card installing the thermal pad will receive EVGA’s full support, the company says.

For our part, we committed to a re-test once EVGA implemented its planned changes, and that’s what you see going live alongside our GeForce GTX 1080 round-up.

In order to address some of the challenges originally posed by EVGA’s card, we drilled two holes into the backplate, right above the two hottest points identified in our previous tests, and cut a circular part off the thermal pads to give our infrared camera a clearer view of the board. Since we are using the same card as before, the results are directly comparable.

Meanwhile, several images have surfaced in forums allegedly showing that the thermal pads used in manufacturing don't completely fill the gap between the front plate and memory modules. Working with enthusiasts online, we analyzed several EVGA cards and couldn’t find evidence of this gap issue, as the picture below shows:

In every case, the pads were placed correctly, even if they weren’t a tight fit. Furthermore, the pictures on the Internet all show the pads sticking to the front cover, and the gap is between the memory module and thermal pad. To better understand what this means, it’s necessary to know a bit about how graphics cards are manufactured.

Thermal pads are almost exclusively glued onto the memory modules first. The front plate is attached later, together with the backplate. Thus, if a thermal pad is stuck to a plate now, at the very least it had to have firmer contact at some earlier point in time.

Our observation is that if the backplate is removed, and then the additional screws holding the front panel are removed as well, that plate is bent in such a way that it leaves a slightly angled gap, which looks very similar to these images. This is also the case if the backplate has been reattached, but the additional screws holding the front panel have not been screwed back in during reassembly.

There could, however, be gaps in the case of a GeForce GTX 1070 with this same type of cooler if the GDDR5 memory modules are a different height than the GDDR5X modules on 1080 cards. And since the thermal pads on the memory modules are kept as thin as possible, that could lead to the reported behavior.

Soon it won’t matter, though. As of mid-November, EVGA will make its thermal pads 0.2 mm thicker, just to be 100% sure there aren’t any issues. This change applies to the pads included in the thermal mod we requested, and those in mass production as well. Furthermore, the newest retail cards will come with the new BIOS versions already installed.

Because we are curious to see how much the new pads affect EVGA’s cards, we’ll test them in three steps. First, we’ll test the pad between the back of the board and the backplate. Second, we’ll test the previously installed pad and another pad between the front plate and heat sink’s cooling fins. Third, we’ll combine all of the aforementioned modifications together with the new BIOS.

Doing so should tell us if it’s really necessary to flash EVGA’s firmware and accept the disadvantage of higher noise levels. If the thermal pads themselves do the trick, perhaps it’s possible to ignore the software side altogether.

EVGA supplies a small plastic bag containing one large thermal pad (for the back) and a smaller one (for the front). They’re complemented by a supply of original EVGA thermal paste, which we’ll use later when it comes time to prepare the heat sink for modifications to the front plate. The backplate, however, is very easy to remove.

According to EVGA, you should install the larger of the two thermal pads as shown below. However, this positioning means that, in some places, the thermal pad will have direct contact with a protective foil (rather than the backplate itself). EVGA uses this foil to seal most of the holes in the backplate. We decided to leave the backplate in its original condition, but recommend removing these thermally unfavorable coverings to create more contact surface and improve airflow.

The darker area on the PCB shows where the thermal pad was positioned during installation. Since the visible area is quite large, it can be assumed that the adhesive thermal pads create good contact. This also speaks to their quality.

The voltage regulators are completely covered, and the RAM is covered at least in part. This result could be optimized further if the thermal pad was positioned about four-fifths of an inch (two centimeters) further to the right, as seen from the back of the PCB. In that case, two rectangular cut-outs need to be made in the top and bottom corners on the right-hand side so that the thermal pad doesn't cover two holes needed for screws.

Applying the thermal pads to the front panel is a bit trickier, since EVGA explicitly recommends covering the long-hole cut-out for the coils, creating a closed surface. Furthermore, this closed surface must also be tall enough to provide sufficient contact with the cooling fins of the main heat sink.

The image taken after re-disassembly shows visible stripe-shaped impressions where this contact happens. We made sure to include the area around the coils, which was one of the hottest parts of the board during our previous measurements. This should significantly reduce the heat propagation inside the board, especially towards memory and GPU.

Measurements During The Gaming Loop

Our metrics are demanding, so the card’s power consumption increases to values that very few games actually reach. It’s reasonably safe to assume that the temperatures we measure are indeed representative of a worst-case scenario.

When the card is idle, the memory and GPU both remain below 104°F (40°C).

The first measurements we’re presenting come from the GPU diode after all three stages of the thermal modification. We’ll compare those readings to our results prior to EVGA’s fix.

The temperature measurements without flashing the BIOS look very similar, which of course is due to the old fan curve. This curve changes significantly in the modified BIOS.

Since those temperatures also have a direct influence on GPU Boost clock rates, we add frequency results to the chart below:

With the original fan curve, GPU Boost frequencies follow the temperatures. Physically modifying the card with thermal pads doesn’t significantly change the GPU temperature or GPU Boost values. Only the new BIOS, with its corresponding higher fan speed and acoustic output, cool the GPU noticeably, enabling more aggressive frequencies.

This wasn’t the focus of our original results, though. The focus there was excessive temperatures measured on completely different parts of the board. To follow up, we need our IR camera.

Original Measurements Without the Modification

To recap, our original round-up results showed the memory modules at their officially specified thermal limits, even during our Metro: Last Light gaming loop.

We measured the card in its original state a second time, with the backplate attached, just to be really sure. The noise levels remained more or less the same.

Measurements After Installing the Thermal Pad on the Back

First, we want to see how much the thermal pad between the circuit board and backplate improves cooling on its own. This may serve as an example in the future, predicting what another poorly-cooled card might gain from similar modifications. The upside here is that we don’t need to remove the cooler; we can perform the mod without handling thermal paste or worrying about parts we took off.

As we can see, memory modules and MOSFETs are about nine degrees Kelvin cooler! However, the memory temperatures are still too high during our stress test. We also note that the GPU package (not the GPU itself) is hotter due to a significantly warmer backplate. Now we know why the temperature of the GPU diode was a little higher with thermal pad in place compared to the original readings without the pad.

It's also noteworthy that the card is a little bit quieter, which could be due to the lower overall GPU diode temperatures, even if the peaks are a little higher at times. On the other hand, 0.5 dB(A) doesn’t make an audible difference:

Measurements with Thermal Pads on Front and Back

We are going to throw in another thermal pad and connect the upper side of the front plate to the large heat sink using thermal paste. This solution clearly improves heat dissipation, especially for the coils and area around the VRMs. A closer look at the measurements shows that the VRMs are already 16 degrees Kelvin cooler than the first test without thermal pads. That’s incredibly significant.  

The memory enjoys plenty of relief as well: almost 15 degrees Kelvin in the gaming loop and at least seven degrees during the stress test. These modifications improve the card's cooling enough to allow some reserves for warmer days. With them, EVGA has no reason to shy away from the competition.

The thermal mod with both pads in place leaves us with similar results; we observe similar noise levels as the original, too.

Measurements of all Modifications, Including the BIOS

Even if it doesn't seem necessary at this point, we still have one more option for improving thermal performance. As such, we flashed the BIOS on our card to EVGA’s latest, available on the company’s website.

The fans spring into action and jump to high RPM, yielding impressive results. The GPU Boost frequency increases, but so does the noise level. Originally, the fan spun at 1300 RPM; now it’s over 1500 RPM. That ~200 RPM jump exacts a notable price that we’ll quantify shortly.

A look at the temperatures makes us wonder why the original BIOS didn’t have the fan spinning 100 RPM faster already. We think the new firmware’s fan curve is a little too aggressive. Even half of the increase would have sufficed.

An incredible 25-degree Kelvin drop on the VRMs in our gaming loop and stress test is a testament to EVGA’s effort. But was it really necessary to go to such extremes?

Now we want to see what these measures do to acoustics. A reading of almost 41 dB(A) the card is distinctly audible and what most would consider loud. The only upside is a more uniform fan noise that can be filtered relatively easily with an insulated case.

A more moderate curve would have left the card between 37 and 38 dB(A), and sufficiently cooled. In the end, the BIOS update gets a small question mark, though we can at least understand it’s easier to implement than a hardware modification.

Conclusion

Two months passed after we first contacted EVGA at the beginning of September and the publication of this update on Tom’s Hardware DE. The company should have probably reacted faster to counter the forum posts from unhappy customers discussing poor thermal performance.

In the end, though, it’s the final result that matters, and EVGA did successfully solve its issue. If the ACX 3.0-equipped cards are used in cases with at least moderate airflow, the BIOS update is superfluous. EVGA would be well-advised to simply install the two thermal pads in mass production and leave the new firmware on its site as an optional download. Otherwise, you’re going to be subjected to 41 dB(A) of noise for a slightly higher GPU Boost frequency. Adding the thermal pads is enough to achieve what gamers want: a cool, but still pleasantly quiet card. Enthusiasts willing to go with the noisier option for a bit of extra performance can flash the BIOS or, even easier, set up a custom fan curve with EVGA’s Those who absolutely want to go with the noisier option to gain a smidgen of extra performance could either just flash the BIOS, or (even easier) set up a custom fan curve with EVGA's own Precision XOC software.

MORE: Best Deals

MORE: Hot Bargains @PurchDeals

Galax/KFA² GTX 1080 Hall of Fame

This card, which is sold in Germany under the name KFA² GeForce GTX 1080 Hall of Fame, is marketed internationally using the Galax brand. Both are exactly the same, though. They come from Galaxy Microsystems, which also refers to itself as Galaxy and Galaxytech. You won't find the company's cards on Newegg or Amazon. Rather, the only way we've found to buy the 1080 HoF is through galax.com.

Why might you feel compelled to poke the company's site? To begin, the 1080 Hall of Fame has a very different look. It sports a white shroud, white body panels, a white backplate, and even a white PCA. Anyone with an MSI Titanium-series motherboard (Z170A or X99) has a match made in heaven.

Beyond the premium paint scheme, we're also impressed by Galax's bundled extras. Among the extraneous gimmicks, you also get a structural support to keep the heavy card from flexing in its slot.

Technical Specifications

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Exterior & Interfaces 

The cooler shroud is made of blindingly white plastic with aluminum highlights for additional eye candy. Overall, Galax went for a mechanical appearance with lots of bold edges and corners.

Weighing 46oz (1315g), this card is very heavy, which explains the support you get in the box. With a length of 12.5in (31.7cm), smaller cases may have a hard time accommodating the 1080 Hall of Fame. A height of five and one-third inches (13.5cm) from the motherboard slot's top edge isn't too subtle either. Although this is a dual-slot card on paper, some smaller parts are an extra 2mm wider, resulting in a total width of 1.46in (3.7cm).

The back of the board is covered by a single-piece aluminum plate with ventilation holes cut into it. Plan for an additional one-fifth of an inch (5mm) in depth beyond the plate, which may become relevant in multi-GPU configurations, particularly if your motherboard's PCIe slots are two spaces apart.

We heavily discourage using this heavy card without its backplate. Despite the front-side mounting and bundled cooling frame, structural stability would suffer significantly without that support.

The top of the card features a back-lit "Hall of Fame" logo and two white auxiliary power connectors.

Vertically-oriented fins prevent air from exiting the front or back of the card. A button on the output bracket lets you set the fans to run a maximum speed; it serves no other purpose and definitely isn't a BIOS switch.

The rear bracket features five outputs, of which a maximum of four can be used simultaneously in a multi-monitor setup. In addition to one dual-link DVI-D connector (be aware that there is no analog signal), the bracket also exposes one HDMI 2.0b and three DisplayPort 1.4-ready outputs. The rest of the plate is mostly solid, with several openings cut into it that look like they're supposed to improve airflow, but don't actually do anything.

Board & Components

As far as memory is concerned, the 1080 HoF is the same as its competition. It uses GDDR5X memory modules from Micron, which are sold along with Nvidia's GPU to board partners. Eight memory chips (MT58K256M32JA-100) transferring at 10 MT/s are attached to a 256-bit interface, allowing for a theoretical bandwidth of 320 GB/s.

The rest of the PCB is considerably more interesting, though. KFA²/Galax proudly advertises a 12+3-phase layout. That's not technically true, so how does the company arrive at the number on its marketing material?

Let's start with a look at the GPU's voltage regulation. Unlike other manufacturers, KFA²/Galax went with International Rectifier's good old IR3595A, which we used to find on a lot of AMD graphics cards. This chip can be used as either a single-loop or dual-loop PWM controller. All six phases provided in loop one are used for the GPU. The rest are regulated by other means.

The reasons why KFA² went with the IR3595A, even though it still uses VRM11 instead of OVR4, are likely found in the fact that it offers firmware programmers more options for customization and complete control over voltage regulation. This also provides some distinct advantages for extreme overclocking, since the chip allows the use of additional extensions. In short, there are benefits. On the other hand, it may conflict with the sensor loops of third-party tools like GPU-Z if monitoring is activated.

But where does the 12-phase marketing claim come from? The 1080 HoF has a small IR3599, which implements a so-called doubler that splits each phase in two individual converter circuits. Thus, there are six true phases and 12 converter circuits.

This isn't a new trick. It does improve the distribution of current, creating a larger cooling area for the MOSFETs. Furthermore, the shunt connection also reduces the circuit's internal resistance as a whole.

Since such a high number of converter rails also leads to a shortage of space, this card uses a highly integrated IR3555 with high- and low-side MOSFETs, a gate driver, and Schottky diode on a single chip, freeing up a little bit of real estate. Spoiler alert: compared to EVGA's GTX 1080 FTW, this concept works very well when it comes to cooling the card.

In order to fit all of the requisite components on its circuit board (that's 12 coils for the GPU and three for the memory/memory controller), Galax relies on vertically-stacked coils encapsulated in a ferrite frame. They're more susceptible to vibration though, and tend to buzz. We'll explore their impact on acoustics shortly.

Three phases are reserved for the GDDR5X and memory controller, employing conventional N-channel MOSFETS for the high- (MDU 1514) and low-sides (MDU 1511), along with an external gate driver. Additionally, a uP9509 is used as the PWM controller. The coils are the same ones used for the GPU.

Current monitoring is handled by an INA3221. Two familiar capacitors are installed right below the GPU to absorb and equalize peaks in voltage.

Power Results

Before we look at power consumption, we should talk about the correlation between GPU Boost frequency and core voltage, which show such striking similarity that we deliberately put their graphs one after the other. KFA²/Galax sets a very high power target that facilitates a relatively constant GPU Boost frequency. It only drops slightly as the temperature increases, and our voltage readings mirror that behavior.

After warming up in a variable-load gaming scenario, the GPU Boost clock rate that started at 2025 MHz stabilizes at 1987 MHz, and then slides to 1923 MHz under constant load. While we measure up to 1.062V at first, voltage later drops to an average of 1.043V.

Summing up measured voltages and currents, we arrive at a total consumption figure we can easily confirm with our test equipment by monitoring the card's power connectors.

As a result of Nvidia's restrictions, manufacturers sacrifice the lowest possible frequency bin in order to gain an extra GPU Boost step. So, the GTX 1080 HoF's power consumption is disproportionately high as it idles at 278 MHz.

Now let's take a more detailed look at power consumption when the card is idle, when it's gaming at 4K, and during our stress test. The graphs show the distribution of load between each voltage and supply rail, providing a bird's eye view of variations and peaks:

Temperature Results

Naturally, heat output is directly related to power consumption, and the 1080 HoF's ability to dissipate that thermal energy can only be understood by looking at its cooling solution.

As with the 1080 Founders Edition, the backplate is mostly aesthetic; it doesn't serve much practical purpose. At best, it helps with the card's structural stability. But the thermal solution is completely sufficient anyway, as our benchmarks show.

KFA² applies a black coating to the back of its plate, improving the absorption of heat. The mounting and cooling frame is made from injection-molded aluminum and placed on top of the PCB. Then it's attached to the backplate with several screws. This structure, which sandwiches the board, also cools the voltage regulators and memory modules through thermal pads.

Why does this concept work so well on the GTX 1080 HoF and not on EVGA's GTX 1080 FTW? Looking at the top of the frame, we see a number of fins that not only increase the cooler's surface area but also take better advantage of available airflow.

The heat sink itself is one of the largest and heaviest you'll find, aside from the solutions that Palit and Gainward use. But KFA²/Galax doesn't stop there.

The cooler relies on a copper block topped with one quarter-inch (6mm) pipe and three more measuring one-third of an inch (8mm) in diameter. During our gaming test, we observed temperatures of only 151°F (66°C) on an open bench, and 154°F (68°C) to 156°F (69°C) in a closed case. Clearly this card's thermal performance is good.

But the high power target does take its toll during our stress test. Power consumption as high as 240W heats the GPU up to 162°F (72°C) on an open bench and 165°F (74°C) in a closed case. Then again, none of us actually play FurMark.

One look at the infrared images reveals the effectiveness of well-planned VRM and memory cooling, even without the need for additional built-in heat sinks. On the other hand, KFA²/Galax's massive mounting/cooling frame wouldn't fit on smaller cards.

A reading of 154°F (68°C) below the armada of VRMs sets a new record, pleasantly surprising us in the process. The GPU enjoys competent cooling as well. In the end, the package is hotter than its graphics processor.

During our stress test, the cooler has to handle an additional 46W, driving temperatures higher. Yet even a reading of 167°F (75°C) is still mild compared to some of the other cards we're benchmarking.

Sound Results

We started looking at Galax's 1080 HoF back in July, and the initial results identified some acoustic issues. Our German team contacted the manufacturer directly and, together with its R&D department, solved the problems we wrote about. The newest cards have been modified according to our guidance, and as of August 2016, they're in mass production. If you buy today, you'll get one of the improved boards.

Let's start with the fan curve, which shows that KFA2/Galax made a conscious decision to eschew semi-passive operation, even at idle. The complaint we originally had was a relatively high 33 percent minimum duty cycle. This is what we saw:

A 33.6 dB(A) result is hardly what you'd call quiet. Our first attempt to tweak the fan controls down to 28 percent duty cycle still left us at 32.6 dB(A). Low-frequency peaks were particularly obtrusive, and the sandwiched PCB caused those sounds to resonate in our chassis.

Under full load, the fans wound up to maximum performance and became even more noticeable. During our stress test they peaked near 38 dB(A). We could have lived with 36.1 dB(A) during our gaming loop, but those low frequencies just dominated the otherwise tolerable sounds.

After the testing we published in July, KFA²/Galax passed our measurements on to its fan supplier with a request for comment and a demand for improvement. When the dust cleared, the manufacturer provided a history of the issue.

In almost every card, regardless of vendor, the fan modules are screwed directly to cover. Depending on the resonance frequency of this structure, there is a chance it'll match the fan noise and form an acoustically disastrous alliance. In order to prevent that, Galax installs those modules without any direct contact.

The fan vendor's measurements not only confirmed our objections, but also proved the effectiveness of its solution. This included a blind test carried out with employees and third-parties.

As the saying goes, trust but always verify. That's why we want to get our hands on current versions of KFA²/Galax GeForce GTX 1080 Hall of Fame and GeForce GTX 1070 Hall of Fame, which would allow us to retake our measurements. We'll publish the results in an update when this happens.

MORE: Best Deals

MORE: Hot Bargains @PurchDeals

Gigabyte GTX 1080 G1 Gaming

Although Gigabyte sells six different versions of the 1080, we're starting with the company's GeForce GTX 1080 G1 Gaming. On the next page we add its 1080 Xtreme Gaming model. Once we make the rounds and cover more of the competition's hardware, we'll circle back around to see what else Gigabyte has to offer.

Let's start with a summary of the 1080 G1 Gaming's technical specifications:

Technical Specifications

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Exterior & Interfaces

Gigabyte's shroud is made of thin plastic and doesn't exude the same high quality as the metallic covers over some of its previous Windforce-equipped models. The trade-off is that this card weighs just 31oz (871g), despite measuring 11 inches (28.4cm) long, 4 1/3 inches (11cm) high, and 1 3/8 inches (3.5cm) wide.

The card is covered by a single-piece backplate. It lacks openings for ventilation, but is connected to the hot spots via thermal tape. Plan to accommodate an additional one-fifth of an inch (5mm) in depth beyond the plate, which may become relevant in multi-GPU configurations.

While it is perfectly possible to use this card without its backplate, removing it requires disassembling the whole cooler and likely voiding Gigabyte's warranty.

The top of the card is dominated by a centered Gigabyte label, glowing in bright colors, and an LED indicator for the card's silent mode. The eight-pin power connector is rotated 180° and positioned at the end of the card. The design may be a matter of taste, but we don't think it's anything to write home about.

At its end, the card is completely closed off. This makes sense, as the fins are positioned vertically and won't allow any airflow down there anyway.

The rear bracket features five outputs, of which a maximum of four can be used simultaneously in a multi-monitor setup. In addition to one dual-link DVI-D connector (be aware that there is no analog signal), the bracket also exposes one HDMI 2.0b and three DisplayPort 1.4-ready outputs. The rest of the plate is mostly solid, with several openings cut into it that look like they're supposed to improve airflow, but don't actually do anything.

Board & Components

Gigabyte's Xtreme Gaming card, dissected on the next page, employs a different layout than the G1 Gaming, which has just enough space for one eight-pin power connector.

The card uses GDDR5X memory modules from Micron, which are sold along with Nvidia's GPU to board partners. Eight memory chips (MT58K256M32JA-100) transferring at 10 MT/s are attached to a 256-bit interface, allowing for a theoretical bandwidth of 320 GB/s.

The 8+2-phase power supply relies on the sparsely documented µP9511P as a PWM controller, just like Nvidia's reference cards. Since this component can't communicate directly with the MOSFETs on the VRM phases, Gigabyte utilizes separate PWM drivers (gate drivers) to talk to a total of three single-channel MOSFETs per phase. We were pleasantly surprised by this rather elaborate solution.

Not using more cost-effective dual-channel MOSFETs (and thereby sacrificing space on the card) may have been a deliberate choice to improve efficiency and dissipate heat more effectively, especially since the third line of MOSFETs is positioned on the back of the board and may benefit from cooling through the board's backplate. This is a unique feature in the field of 1080s we're reviewing.

As seen previously, two capacitors are installed right below the GPU to absorb and equalize voltage peaks. The board layout seems pretty dense, but it's organized well.

Power Results

Before we look at power consumption, we should talk about the correlation between GPU Boost frequency and core voltage, which show such striking similarity that we deliberately put their graphs one after the other. Notice how the two curves don't drop as much in the face of rising temperatures compared to Nvidia's reference design.

Where the GPU Boost frequency falls to 1873 MHz under load (and even further during our stress test), this movement is mirrored by our voltage measurements. We recorded up to 1.062V in the beginning (similar to the Founders Edition model), and the value drops below 0.962V later on (still above the readings collected for our Founders Edition sample).

Summing up measured voltages and currents, we then arrive at a total consumption figure we can easily confirm with our test equipment by monitoring the card's power connectors.

As a result of Nvidia's restrictions, manufacturers sacrifice the lowest possible frequency bin in order to gain an extra GPU Boost step. So, Gigabyte's power consumption is disproportionately high when idle. In all fairness, the company manages this behavior relatively well compared to some of its competition.

Our complete power measurements are as follows:

Now let's take a more detailed look at power consumption when the card is idle, when it's gaming at 4K, and during our stress test. The graphs show the distribution of load between each voltage and supply rail, providing a bird's eye view of variations and peaks:

Temperature Results

Naturally, heat output is directly related to power consumption, and the 1080 G1 Gaming's ability to dissipate that thermal energy can only be understood by looking at its cooling solution.

Gigabyte relies on its aging Windforce design. In this case, the copper heat sink gives way to a simpler (and more affordable) direct heat touch solution that facilitates contact between the cooler's three partially flattened pipes made of sintered composite material and the GPU.

The rest of the plate is meant to cool the memory and (commendably) the voltage regulators. Added thermal pads ensure good contact.

The slightly better performance of this cooler compared to Nvidia's Founders Edition card is also reflected by the temperature graph, as a limit of 158°F (70°C) is not, or hardly (162°F/72°C in a closed case) exceeded.

Unfortunately, upon a closer look at the infrared images, it becomes clear that this card does have troublesome areas. While the voltage regulation circuitry stays relatively cool at 180°F (82°C), two smaller SMD capacitors glow at an unpleasant 189°F (87°C).

During our torture loop, those capacitors (R990 and R991) even reach temperatures of 212°F (100°C), despite a marginal increase in heat measured at the voltage regulators.

To test the two shunts more accurately, we change the distance to just 2.75in (7cm) and swap the IR camera's lens, since the areas we need to measure are just 1mm². A look at the specs and an email to Gigabyte quickly confirmed that our findings shouldn't cause any problems.

Even after a FurMark run, the measured temperature proves acceptable if the specifications are to be trusted. Of course, that's not taking into account the PWM controller residing right above the two capacitors on the board's top side. It generates a fair amount of heat as well, resulting in two hot spots right on top of one another.

Our initial testing showed that the severity of this problem could easily be mitigated with an additional thermal pad, so Gigabyte implemented our suggestion on its retail cards.

With the backplate mounted (sporting thermal pads between the RAM modules and VRM), the "outside" temperatures appear as follows:

Sound Results

Since power consumption during our gaming and stress tests is similar, it comes as little surprise that the fan curves are also a close match. The behavior during start-up and an overall clean profile reliably prevents multiple on/off cycles once the fans spin up. In addition, the start-up speed is chosen in such a way that the fans should continue to start reliably even as they age. The same goes for turning the fans off once the card cools down.

When the card is idle, its noise is not measurable thanks to a semi-passive mode that keeps the fans from spinning. However, running at full load for a while easily raises the noise level to 40 dB(A). It gets even louder during our torture test.

To be frank, we're not impressed with the card's acoustics. They can be improved by manually adjusting the fan curve and increasing the temperature target to 167°F (75°C). Doing so reduces the maximum GPU Boost frequency though, and you may start to hear an audible chirp coming from the coils.

Considering what other manufacturers achieve using aggressive coolers, the G1 Gaming's solution almost feels unambitious. It's missing the tender loving care competing vendors put into their own 1080-based cards. Though the G1 Gaming isn't bad in this regard, the end result isn't particularly good either.

MORE: Best Deals

MORE: Hot Bargains @PurchDeals

Gigabyte GTX 1080 Xtreme Gaming

It took a while for us to get our hands on Gigabyte's GeForce GTX 1080 Xtreme Gaming, but without giving too much away, the wait was worthwhile.

Gigabyte advertises this card's extended feature set, so you know we have to pay particular attention to its value-adds. Upon opening the over-sized box, we discover loads of useful extras.

While wrist protection and a new mouse pad are nice to have (even if they're more Gigabyte swag than essentials), the VR bracket does deserve special mention. That's because the GeForce GTX 1080 Xtreme Gaming comes with a front panel bracket that exposes USB 3.0 I/O and HDMI outputs for easy-to-reach connectivity.

Of course we'll cover the stacked fans and heat pipes, but let's start with a word on the 1080 Xtreme Gaming's overall performance and how we tested it.

We try to benchmark every card as it ships, right out of the box. That means we typically avoid installing vendor-specific software and intentionally use default firmware settings. Many utilities are only available for Windows, and direct comparisons are complicated by the sheer number of options available for each card. Nevertheless, in this case we're including measurements for Gigabyte's optional OC mode.

The company's four-year warranty is exemplary, even if it requires registering the 1080 Xtreme Gaming online.

Technical Specifications

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Exterior & Interfaces