Back in 2023. a few months after the RTX 40 series had launched, speculation over an RTX Titan coming out in that generation was sky-high. Ultimately, it was never released, which only built up the myths of this GPU that has since slowly become less and less mysterious through prototype leaks. One of those prototypes is in the hands of community legend Roman 'der8auer' Hartung, who's benchmarked it previously, but he's now finally taken it apart to show us what's inside.

We first see how the card is built from the outside; it's similar to a Founder's Edition RTX 5090 where the entire shroud looks like one big heatsink. When looking at it from the top, you can see through the card entirely since there's no PCB or anything else in the way like in a conventional GPU. There's a fan on either side of the heatsink, but also a smaller third fan in the middle that's only visible at certain angles.

Der8auer remarks how unique this design is, and would've been for its time had it actually launched before the RTX 5090, like it was clearly planned to at one point. The card is gigantic, feeling almost comically large in our host's hands, since it's a monstrous quad-slot design.

This can be confirmed by just looking toward the back, which houses the vertically stacked I/O in the form of 3x Display Port 1.4a and 1x HDMI 2.1 — these are actually soldered onto the mainboard as you'll see later, unlike how the RTX 5090 FE handles them via a daughterboard. Anyhow, unscrewing the I/O shield here is the first step of getting inside the card.

Once removed, you can see a parting line where der8auer wedges a screwdriver that separates the side plate previously held by the I/O shield. The side plate eventually comes off, and it has heatpipes running through the back with thermal pads placed to make contact with components on the PCB like the memory and VRM controller.

That brings us to the main attraction, the actual PCB, which is mounted on the side of the card, similar to the RTX 4090 Ti prototype we looked at a while ago. The PCB has 12x 2GB GDDR7 memory modules on the top side that's visible right now, with 12x more on the back, for a combined total of 48GB VRAM.

Now, to get the PCB out, the center frame of the GPU needs to be removed, so the card is flipped over and the large silver "X" plastered across the chassis is carefully taken apart. Underneath this are channels hiding the fan ribbon cables that also need to be plugged out from both ends: one in the center of the top frame, and the other connected to the side of the main PCB.

Lastly, there are some more screws on the mainboard, eventually leading to the spring mechanism in the center, which was taken apart off-camera to conceal part numbers to protect the source. The PCB is finally separated from the rest of the chassis, and it's a dense piece of engineering marvel that will impress even Apple engineers, with the AD102 GPU gloriously sitting in the middle.

Unlike a traditional GPU that has the PCIe x16 connector as part of the mainboard, we can see ribbon cables emanating from one side holding a daughterboard for the PCIe, sort of like how a riser cable looks in a vertical GPU mount. Toward the right of the board are the aforementioned display ports, but toward the left are three contact points for the power supply and these are really cool.

There are contact points built into the cooler of this Titan Ada prototype that touch these contact points and provide power. Der8auer says one of the points is likely for ground while the other two are for each of the 12VHPWR connections. Above those dangle the sense pins that were disconnected from the PCB as well.

The final step remaining is to separate the heatsink from the aluminum frame since everything has been removed, which slides off nicely because it's essentially friction-fit at this point. Now that we get a clear look at all the innards, der8auer shows us all the cables running around the perimeter of this chassis, hidden by endless, segregated panels and side pieces that were taken apart till now.

With the entire graphics card disassembled, der8auer said it took him one and a half hours to put it back together, and he had to scrub through footage just to make sure he didn't miss a screw here or there. He once again commends Nvidia's engineering efforts, saying the RTX 5090 and this RTX Titan Ada prototype are the most impressive GPUs he's ever seen.

Many of the best graphics cards currently available utilize the 16-pin (12VHPWR) power connector. Due to several unfortunate incidents involving meltdowns of this connector, numerous manufacturers have introduced products designed to prevent such occurrences. German liquid-cooling specialist Aqua Computer is the most recent brand to enter this market with Ampinel. However, what truly distinguishes Ampinel from other solutions is that it incorporates active current balancing.

The 16-pin power connector meltdowns have been claiming victims since the GeForce RTX 40 series (codenamed Ada Lovelace). A long-standing theory regarding these misfortunes suggests that Nvidia removed load-balancing circuitry from its Ada Lovelace graphics cards, a feature that was present in the previous GeForce RTX 30 series (codenamed Ampere). This omission is one reason why cases of the 16-pin power connector melting have not been reported on GeForce RTX 3090 Ti graphics cards. Many industry observers speculate that the design decision was made to reduce production costs.

To date, most, if not all, devices related to the safety of the 16-pin power connector have adopted a passive approach concerning the meltdown issue. These devices primarily monitor either the current or the temperature of the 16-pin power connector. Conversely, Ampinel adopts a more proactive strategy by implementing a load balancer to evenly distribute loads across the various wires within the 16-pin power cable.

The Ampinel features a six-channel load balancer that utilizes a microcontroller to continuously monitor the six 12V power lines inside the 16-pin power connector and regulate the current flow in real-time. Upon detecting that the current in any power line exceeds 7.5A, which is the rated current per contact, Ampinel intervenes to redistribute the load to prevent overheating that could potentially provoke a meltdown of the power connector.

The Ampinel features a compact 128 x 64 resolution OLED display, which outputs the current for each of the six 12V lines. Additionally, it includes an integrated buzzer, capable of emitting a loud sound (approximately 85 dB), designed to provide an audible warning when any of the power lines deviates from specifications. While the RGB lighting may appear aesthetically pleasing, it primarily functions as a visual alert system. Imagine the Ampinel sounding like a siren from an ambulance.

You may refine the Ampinel device using the included Aquasuite software. The manufacturer emphasizes that the device's emergency shutdowns, alarm outputs, and visual and acoustic alarms operate independently of software or a USB connection. However, it is necessary to utilize the software to customize and permanently save your personalized presets on the Ampinel or enable monitoring in Windows.

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB Introduction

The Nvidia GeForce RTX 5060 Ti 16GB and RTX 4060 Ti 16GB represent the new and old guards of the 16GB realm for Team Green. Both of these are mainstream GPU offerings, ostensibly priced in the $430–$500 range. We weren't particularly impressed with the 4060 Ti 16GB at launch, but our feelings have tempered over time — helped by an unofficial $50 price cut — with both having appeared on our list of the best graphics cards.

The Nvidia Ada Lovelace 4060 Ti 16GB card originally launched over a month after the initial RTX 4060 Ti 8GB card, with a $100 price premium. Throughout the ensuing two years, the 8GB card has been more readily available, though as noted already, street prices on the 16GB card did drop by $50 at times. 

Conversely, the RTX 5060 Ti 16GB and 8GB variants technically launched at the same time on April 16, 2025, with only a $50 price gap. So far, the 16GB models have been far more widespread in the U.S., to the point where we have not yet been able to purchase an 8GB card at retail for testing.

Given the ongoing concerns with having only 8GB of VRAM, we feel it makes the most sense to compare the new and old 16GB xx60 Ti cards. We'll look primarily at performance, as well as power, features, and pricing. In theory, this should be a very easy win for the new Nvidia Blackwell GPU, but let's see how they actually stack up.

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB Specifications

Specifications for the RTX 5060 Ti 16GB and RTX 4060 Ti 16GB are mostly similar, with a few key upgrades for the newer GPU. While both the GB206 and AD106 dies have a maximum of 36 SMs (Streaming Multiprocessors), the 5060 Ti uses the fully enabled chip while the 4060 Ti disables a pair of SMs. That cascades through the GPU shaders, tensor cores, RT cores, and TMUs, giving the 5060 Ti a theoretical 6% advantage. However, clock speeds also factor into performance, and the 5060 Ti comes with a 2,572 MHz boost clock compared to 2,535 MHz on the 4060 Ti, another 1.5% advantage.

The core counts are a "hard" spec, but the clock speeds are a bit fuzzy. Power limits also come into play, and the 5060 Ti has a 180W TGP compared to 165W on the 4060 Ti 16GB. In practice, across our 21-game test suite, the 5060 Ti 16GB averaged clock speeds of 2,776 MHz compared to 2,725 MHz on the 4060 Ti 16GB — a slightly larger 1.9% advantage in favor of the newer card. Combined with SM and core counts, that gives the 5060 Ti 16GB an 8% advantage in raw compute.

That's pretty negligible in terms of a gen-on-gen increase, but architectural updates can further impact the real-world performance. More pertinent than the raw compute is the memory subsystem. The RTX 5060 Ti 16GB has 28 Gbps GDDR7 memory while the RTX 4060 Ti 16GB has 18 Gbps GDDR6 memory. That means the newer part has a 56% advantage in raw memory bandwidth — though architecture and timings can still play a role.

Taken together, based on the specifications, you should expect to see the newer GPU outperform its predecessor by anywhere from 8% to 56%, depending on whether a particular game or workload relies more on raw compute, memory bandwidth, or a blend of the two. In practice, compute tends to be the more important metric. As such, we'd expect a 15~25 percent lead from the 5060 Ti. Let's see how things go in actual benchmarks.

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB Gaming Performance

For most people, gaming performance will be the most important factor in choosing a graphics card. We have a test suite of 21 games, 6 of which have ray tracing enabled and 15 with pure rasterization. Upscaling and frame generation are disabled for all these tests, as we view those as "performance enhancers" rather than baseline features that should be compared.

Both the RTX 5060 Ti 16GB and RTX 4060 Ti 16GB support Nvidia's various DLSS features, though the 5060 Ti also offers MFG (Multi Frame Generation) — which can smooth out performance but definitely doesn't improve the actual feel of games that support it as much as the numbers might suggest. The experience of DLSS 3 framegen in our view tends to be inflated by about 50–60 percent (meaning, divide the resulting number by 1.5–1.6 to get a realistic view of what it feels like), while MFG 4X is about a 170–200 percent inflation (divide by 2.7–3.0).

Those are fuzzy approximations, but basically if you get less than a 50% "increase" in framerates via framegen, we would argue that it actually feels like a step backward. Likewise, less than a 170% increase from MFG 4X would be a net loss. Or to put hard numbers on it, if the base performance is 60 FPS, framegen needs to run at 90 FPS or more and MFG 4X needs to run at 160 FPS or more. Some would argue even those figures are too generous, but our main point is that "FPS" inflation via frame generation techniques has been heavily abused as a true indication of performance.

The charts and tables provide all the important numbers, so we're not going to dig into every single result. We're mostly interested in the overall trends, recognizing that individual games can come in above or below the averages. That's why we have an extensive 21 game test suite, with equal weighting given to each result via a geometric mean calculation.

The nice thing about comparing two Nvidia GPUs, even though they're different architectures, is that relative performance ends up being pretty consistent across both rasterization and ray tracing games. AMD and Intel GPUs vary quite a bit more, depending on what specific games are used.

Overall, the 5060 Ti 16GB outperforms the RTX 4060 Ti 16GB by 16% at 1080p medium, 17% in the ray tracing suite, and 16% in the rasterization suite. That means the 5060 Ti and 4060 Ti generally aren't CPU limited, even at 1080p medium. The minimum FPS (1% lows) shows similar consistency. Looking at the individual games, the 5060 Ti 16GB leads by 6% (Assassin's Creed Mirage) to as much as 29% (A Plague Tale: Requiem).

Bumping up to 1080p ultra settings, most of the deltas are virtually unchanged. The 5060 Ti 16GB still leads by 16% overall, 17% in the ray tracing suite, and 16% in the rasterization games. Minimum FPS ends up slightly closer overall at 18%, but with a wider 23% spread in our ray tracing tests and 16% in the rasterization games. Across all 21 games, we measured an 8% (Starfield) to 26% (Plague Tale: Requiem and Space Marine 2) delta in favor of the 5060 Ti.

Given the memory bandwidth advantage, the 5060 Ti 16GB should increase its overall lead at higher resolutions, and it does — just not by much. It's 18% faster overall at 1440p ultra, and 18% in both the rasterization and ray tracing suites. The various games show an advantage of 9% (Starfield) to 29% (Plague Tale and Space Marine 2, again).

Finally, at 4K ultra, we see the biggest lead for the 5060 Ti 16GB. Across the full 21-game suite it's 21% faster than the 4060 Ti 16GB, with a 20% lead in the ray tracing games and a 22% lead in the rasterization tests. The smallest advantage improves to 12% (Avatar), with the biggest lead being 33% (Space Marine 2).

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB AI and Content Creation Performance

Looking at other potential workloads, we have various disciplines that all loosely fall under the content creation umbrella. AI workloads include Stable Diffusion image generation, LLaMa text generation, image classification, inference, and other tasks. We also have 3D rendering via Blender, video transcoding using SPEC Workstation 4.0, and the Viewport test suite from SPEC Workstation 4.0.

Many people — particularly gamers — may never do any of these tasks on their PC, though it's worth nothing that Nvidia's push for "neural rendering" techniques can also extend into AI-powered NPCs. That's not something we can directly benchmark at present, but there's potential for the AI performance to become more important in the coming years if AI NPCs catch on in some major games.

Of all the content creation tasks, we think the AI tests deserve the most attention. As noted above, such workloads could become a bigger part of how people use their PCs and graphics cards in the future, even within games. Overall, the 5060 Ti 16GB delivers about a 24% lead in AI workloads compared to the RTX 4060 Ti 16GB. It's also worth mentioning that none of these AI tests at present leverage the FP4 support that was added to the tensor cores in the Blackwell architecture, so there's potential for even more of an AI deficit if future applications make use of that number format.

Video transcoding ends up being a virtual tie, indicating there's not a huge difference between the NVENC/NVDEC blocks in Ada and Blackwell. The 5060 Ti takes a 4% lead by virtue of running at slightly higher clock speeds, though Nvidia did supposedly improve the encoding quality — that's not something these tests measure, unfortunately.

3D modeling using Blender basically matches what we saw in gaming performance. Overall, the 5060 Ti 16GB runs 20% faster than the 4060 Ti 16GB, with a spread of 14% to 29% in the three different scenes. SolidWorks agrees with that number, while 3ds Max shows an even larger 31% lead for the 5060 Ti.

Those last two are part of the SPEC Workstation 4.0 Viewport tests, which show a 17% lead overall for the 5060 Ti, with a range of 1% (Creo) to as much as 31% (3ds Max). Most of the SPEC Workstation tests are truly for professionals, however, so we wouldn't put too much weight on them. Most paid professionals will have a workstation equipped with a workstation GPU (e.g. Nvidia RTX 6000 Ada Generation or the new RTX Pro 6000 Blackwell).

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB Features and Software

The vast majority of Nvidia features and software tools are available on both the RTX 4060 Ti 16GB and the RTX 5060 Ti 16GB. As such, it's easy to think that the two are "tied" in features. However, there are a few noteworthy exceptions.

First, the Blackwell RTX architecture adds support for Multi Frame Generation. We don't love the way the numbers are used to indicate massive performance improvements (i.e. Nvidia CEO Jensen Huang tried to claim that an "RTX 5070 is as far as an RTX 4090" at the launch event), but MFG isn't inherently bad. There are games where it works better, and the additional frame smoothing can make games feel better, subjectively. It's a nuanced discussion, in other words, and not the "fake frames BAD" mentality that often gets tossed around.

Blackwell RTX, and by extension the RTX 5060 Ti 16GB, is also supposed to have better support for so-called neural rendering (i.e. Microsoft's Cooperative Vectors API). We don't yet know precisely how much better it might be compared to older GPUs, but we do know Cooperative Vectors is supposed to work with all RTX GPUs — and that it's supposed to be better on Blackwell RTX.

As noted earlier, the RTX 5060 Ti also has support for FP4 (and FP6) numerical formats in its tensor cores. Making proper use of FP4 presents challenges for AI models, but various algorithms have demonstrated superior performance and reduced memory requirements with a "negligible" drop in overall inference quality. As AI continues to evolve, it's reasonable to expect the use of FP4 to become more commonplace.

We've seen that with FP8 support as well. First introduced with the RTX 40-series GPUs, the use of FP8 modes was relatively rare for most of the past 2.5 years. However, Nvidia's new DLSS Transformers algorithms for upscaling (super resolution) and ray reconstruction appear to leverage native FP8 support. It's why the performance hit for DLSS Transformers tends to be substantially lower on RTX 40- and 50-series GPUs compared to RTX 30- and 20-series GPUs.

But Nvidia giveth and Nvidia also taketh away. Blackwell graphics cards no longer have PhysX support, so the RTX 5060 Ti 16GB simply can't run PhysX in older games that used the feature. (Not many recent games have used GPU PhysX, so this mostly applies to games like the Batman: Arkham series and Mirror's Edge, all of which are at least 10 years old now.) 32-bit CUDA support was likewise retired, which again mostly impacts older apps that are no longer being actively updated.

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB Pricing

The RTX 4060 Ti 16GB launched in July 2023 with an MSRP of $499, $100 more than the 8GB variant. The RTX 5060 Ti 16GB came out in April 2025 with a $429 MSRP. On paper, that clearly suggests a pricing advantage for the 5060 Ti 16GB. In practice, it's also an advantage when shopping around for new cards, but that's because the 4060 Ti 16GB (and the rest of the RTX 40-series) has been retired.

Since both GPU lines are made using TSMC's 4N process node, there's no reason for Nvidia to continue producing the older chips, and in fact all indications are that wafer orders for Ada Lovelace may have ended in the fall of 2024. Supplies have completely dried up and prices have shot up as a result.

The availability of RTX 50-series GPUs has been relatively poor compared to what we saw with the 40-series launch. Sure, the RTX 4090 sold out for the first few months, but eventually it reached the point where you could expect to find some models selling slightly below the base MSRP, and all of the other 40-series GPUs eventually dropped below the base MSRP by late 2023 and early 2024. So far, the RTX 50-series has rarely even approached MSRP, often selling at 20% to 50% above the minimum suggested price.

At present, the least expensive RTX 5060 Ti 16GB card we can find is an Asus Dual RTX 5060 Ti 16GB for $492.98 at Newegg. That's about 15% above the base MSRP that Nvidia set, but Nvidia can't force its add-in board (AIB) partners to actually sell cards at MSRP. With ongoing GPU shortages, it's little surprise that nearly all 50-series GPUs are "overpriced" relative to MSRP.

As bad as that might seem, the RTX 4060 Ti 16GB is in an even worse state. With production of those GPUs likely halted last year, only limited supplies remain. Mostly, it's third-party marketplace resellers on Newegg, Amazon, and similar storefronts trying to make some serious money. Even the 8GB cards start at $575 (MSRP is supposed to be $400), while the cheapest RTX 4060 Ti 16GB costs $679. Yeah, that's a bad choice unless for some reason you specifically need the older hardware.

RTX 5060 Ti 16GB vs RTX 4060 Ti 16GB Final Verdict

This was pretty much a foregone conclusion, but the RTX 5060 Ti 16GB easily beats the RTX 4060 Ti 16GB in every conceivable metric. There's only one potential advantage to the 4060 Ti, and it's nebulous at best: It supports PhysX and 32-bit CUDA (though both are likely deprecated and will fade away in the coming months).

In contrast, the 5060 Ti 16GB offers about 20% higher gaming performance on average, across nearly all tested resolutions and settings. It's also about 20% faster in AI workloads and 3D rendering, and about 17% faster in a selections of professional applications. That's largely thanks to having 56% more memory bandwidth, courtesy of GDDR7, though other architectural upgrades and slightly increased core counts also factor into things.

The real kicker is the price. Unless you're lucky enough to find an RTX 4060 Ti 16GB on a steep discount, the 5060 Ti 16GB costs quite a bit less at popular online retail sites like Newegg and Amazon. Just say no, in other words. Slower performance, missing new features, and a price that's currently 38% higher? Pass. It's also a poor proposition when you look at total system cost (using $750 for the price of the rest of the PC). The 5060 Ti 16GB delivers 24% more performance per dollar than the RTX 4060 Ti 16GB in that case (using MSRPs for both, which benefits the 4060 Ti).

If you're wondering about power use and efficiency, even though the TGP (Total Graphics Power) on the 5060 Ti 16GB is 15W higher on paper, in practice it averaged lower power use in our testing. Power use was 1W higher at 4K ultra (a setting both GPUs struggle with), 5W lower at 1440p ultra, and 8–9 watts lower at 1080p. With 21% higher performance at 4K, that works out to much improved overall efficiency.

Hopefully it goes without saying, but just because the RTX 5060 Ti 16GB offers more performance and costs less than a new 4060 Ti 16GB, that doesn't mean people who already own the prior generation card should be in a rush to upgrade. A 20% performance uplift combined with a few new features might be nice to have, but we generally only recommend upgrading if you can get at least a 50% performance improvement, and doubling performance is often possible if you're upgrading from a card that's a couple of generations old.

Let's also not forget that there are 8GB variants of both GPUs. Again, just say no to such cards. Unless you can score a particularly impressive deal, the lack of VRAM is increasingly becoming a problem for 8GB cards. Sure, older and lighter games can run just fine on such GPUs, but it's the newer games that need faster hardware, and they're having problems. The 4060 Ti 16GB and 8GB offer nearly identical performance in our test suite at 1080p, but the 16GB card pulls ahead by 7% at 1440p, and by 36% at 4K — with the most demanding newer games exceeding those averages. We anticipate the story will be largely the same with the 5060 Ti 8GB.

So there you have it: While the generational gains of the RTX 5060 Ti 16GB over the RTX 4060 Ti 16GB aren't as large as we've seen with prior architectural upgrades, the lack of progress on process nodes and the relatively similar die sizes and transistor counts mean we shouldn't expect 50% performance improvements at the same price going forward. 20% more performance for an ostensibly 14% lower price is about as good as we're likely to see. That's about 40% more performance per dollar, which beats any of the other RTX 50-series GPUs.

RTX 5070 vs RTX 5060 Ti Introduction

The Nvidia GeForce RTX 5070 and RTX 5060 Ti 16GB are the two most recent GPUs launched by Team Green. Since the beginning of 2025, Nvidia has released five Blackwell RTX 50-series GPUs — six if you want to count the RTX 5060 Ti 8GB as a separate entry.

To say that supplies have been limited and insufficient to keep up with demand would be a gross understatement, but the same applies to any of the best graphics cards right now. In fact, of all the Blackwell RTX GPUs, the 5070 and 5060 Ti are currently the most readily available; just don't expect to find most models at MSRP.

Our GPU benchmarks hierarchy ranks all the graphics cards by performance, and naturally the more expensive cards come out ahead of their less expensive siblings. But how do the RTX 5070 and RTX 5060 Ti 16GB stack up? We'll look at the performance, approximate pricing — because no GPU prices are reliably set in stone right now — as well as other aspects of the cards to help you decide which one might be right for your gaming PC.

Most of the features between the two GPUs will be identical. Both use the Blackwell architecture and support DLSS 4 and MFG, for example. Still, let's start with the specifications to see how the 5070 and 5060 Ti stack up.

RTX 5070 vs RTX 5060 Ti Specifications

As you'd expect, the RTX 5070 offers more raw performance on paper and in practice. It uses a GPU that's 45% larger, with 41.6% more transistors. More importantly perhaps, it has 33% more SMs, tensor cores, RT cores, and related computational units. It also has 50% more memory bandwidth and a 50% wider memory interface, plus 50% more L2 cache.

Memory interfaces don't scale as well as core logic on modern process nodes, and there are other aspects of the Blackwell GPU — the video encoders/decoders, display outputs, etc. — that don't really change, which explains why the chip size and transistor counts don't necessarily scale linearly.

Clock speeds aren't quite the same, at least on paper. The 5060 Ti has a 2572 MHz boost clock and the 5070 has a slightly lower 2512 MHz boost clock. But we're also looking at the 5070 Founders Edition with reference clocks, and a PNY 5060 Ti 16GB OC that has a 2692 MHz boost clock. Plugging in those numbers gives the 5070 30.9 teraflops of FP32 compute and the PNY 5060 Ti has 24.8 TFLOPS. That means the 5070 'only' has 25% more compute on paper — which isn't actually all that accurate.

In our full suite of gaming benchmarks, the 5070 Founders Edition averaged clock speeds of 2826 MHz while the PNY 5060 Ti 16GB OC averaged 2776 MHz. That means despite theoretically 180 MHz lower clocks, in practice Nvidia's 5070 card came out 50 MHz ahead.

That means the 5070 FE offers about 34.7 TFLOPS compared to 25.6 TFLOPS, a 35.5% advantage. Practically speaking, though, the 5070 should be up to 35% faster in games that are compute bound, and up to 50% faster for games that are memory bandwidth bound — and potentially the 5060 Ti could close the gap if VRAM capacity comes into play.

RTX 5070 vs RTX 5060 Ti Gaming Performance

There are multiple facets of performance to discuss. We'll start with gaming performance, as that's what the majority of people looking at the RTX 5070 and RTX 5060 Ti 16GB care about the most. We have an expanded test suite of 21 games, with 15 rasterization benchmarks and six ray tracing benchmarks — slightly more than the 18 games we use in our GPU hierarchy (we're planning to add these into the hierarchy but haven't tested every GPU yet).

The charts group things together by overall performance (geometric mean across all 21 games, where each gets equal weighting), rasterization-only performance, and ray tracing-only performance. We also have separate tables that show the percentage differences below the charts (sorted in reverse order due to the vagaries of Excel).

It's safe to say that neither the RTX 5070 nor the RTX 5060 Ti 16GB are primarily targeted at native 4K ultra gaming. The 1440p result works as a proxy for 4K with DLSS quality mode upscaling, and the 1080p result stands in for 4K with performance mode upscaling, though there would be potential differences. Upscaling does have some overhead, but since we're looking at two Nvidia GPUs we can expect a pretty consistent delta cause by DLSS. CPU bottlenecks would potentially affect performance at lower resolutions as well, which we can see by the results.

At 4K, the RTX 5070 leads the RTX 5060 Ti 16GB by 32% on average, with a performance difference of -5% (Indiana Jones and the Great Circle) to +43% (A Plague Tale: Requiem). Indiana Jones is known to want a lot of VRAM — you can even run the ultra setting on cards with less than 12GB — while various other games appear to want more memory bandwidth.

Dropping to 1440p, the 5070 lead shrinks fractionally to 31% overall, and this time there are no performance deficits. There may be a few other games out there that can push beyond 12GB of VRAM use, but from our test suite — a rather demanding suite overall, we'd say — 12GB remains sufficient in all games for 1440p, and in all but one of the games for 4K.

1080p starts to run into CPU limits, even more so at medium settings, which shrinks the delta between the 5070 and 5060 Ti to 28% at ultra settings and 24% at medium settings. That's still a sizeable gap, with a few of the games showing basically equivalent performance due to the CPU bottleneck (Baldur's Gate 3, Flight Simulator 2020, Spider-Man 2, Spider-Man: Miles Morales, and Stalker 2 end up tied in one or both cases).

Looking at the rasterization and ray tracing performance, overall there's not much difference at the various resolutions and settings. 4K ultra does run into VRAM limits on one of the RT games, so that the gap is slightly narrower there than at 1440p and 1080p, and our RT test suite does end up with lower FPS than the rasterization games at ultra settings. Medium settings with RT enabled actually runs faster on average for our test games, but only Control and Cyberpunk are truly demanding RT games in our title selection.

RTX 5070 vs RTX 5060 Ti Content Creation Performance

Our content creation test suite consists of a variety of test that all loosely fall into the "professional" category. We have multiple AI tests, 3D rendering benchmarks, video transcoding performance courtesy of SPEC Workstation 4.0, and the SPEC Workstation 4.0 Viewport benchmarks (basically the same tests as SPECviewperf).

As expected, the RTX 5070 generally walks away with the content creation crown. It's 45% faster in the Blender 3D rendering benchmarks, 33% faster in MLPerf text generation tokens per second, and 20–43 percent faster in the Procyon and SPEC AI tests that we ran. The 5060 Ti 16GB did have a faster time to first token in MLPerf, though it's possible that's due to updates to the benchmark (we'll recheck the 5070 in the near future, once our test rig isn't busy).

For SPEC Workstation 4.0's Viewport tests, there's another instance of the 5060 Ti 16GB garnering a win — it's 2% faster in Creo. But overall the 5070 still has a 23% lead. Handbrake video transcoding meanwhile ends up as a tie, which is expected as the fixed function encoding and decoding hardware is the same, so the only difference would be GPU clocks during that test.

Neither GPU is specifically marketed as a professional solution, though we anticipate there will be professional variants of both chips (with more VRAM and drivers that enable additional accelerations in some of the SPEC Viewport tests). Mostly, the AI results are interesting as a hobbyist solution, though it's possible future games might leverage the AI hardware more.

RTX 5070 vs RTX 5060 Ti Conclusion

Which is better: RTX 5070 or RTX 5060 Ti 16GB? Price, as we said at the start, will arguably be the biggest factor in deciding between the RTX 5070 and RTX 5060 Ti 16GB. It's also the one thing that we can't really pin down at present. Since the 5060 Ti launched just a few days ago, pricing and availability are more likely to fluctuate in the near term. The 5070 has been out for a month now, and we have a better idea of what to expect.

First, let's talk about MSRPs and performance. On paper, Nvidia gives the 5060 Ti 16GB a $429.99 MSRP and the 5070 has a $549.99 MSRP — so $430 and $550. That means the 5070 is supposed to cost 28% more than the 5060 Ti 16GB. And based on the performance results, that's exactly in line with what you get.

Linear performance scaling generally means you're better off buying the more expensive card — we would normally expect to see diminishing returns. So if you buy an RTX 5060 Ti 16GB at MSRP and get 0.171 FPS/$ across our gaming tests, versus the RTX 5070 with 0.170 FPS/$, we give the win to the 5070. The reason is because the GPU doesn't exist in a vacuum; there's the rest of the PC to consider.

Alternatively, let's talk current street prices. These will change dramatically over time, but we can currently find an RTX 5060 Ti 16GB for $509.98, or an RTX 5070 for $619.98. That's the second pair of tables in the above gallery, which obviously skews in favor of the 5070 at current prices. And that's what you need to look at, first and foremost.

Finally, imagine a PC where the CPU, motherboard, SSD, RAM, PSU, and case together cost $750 for example. Now you're looking at a total cost of $1,180 for the 5060 Ti 16GB compared to $1,300 for the 5070 build (using the base MSRPs for the GPUs). That works out to just 10% more total money for about the same 25~30 percent performance uplift. It's what the last set of tables above show, and again, it favors the 5070 heavily.

Given the 25~30 percent performance advantage of the RTX 5070 over the RTX 5060 Ti 16GB, spending up to 30% more on the RTX 5070 would make sense. It's the clear winner at current online prices, or at the given MSRPs. In fact, we would argue that it's worth spending up to 40% more for the 5070 compared to the 5060 Ti, as we normally expect diminishing returns.

We don't know where prices will end up, in the U.S. or elsewhere, but you can use the above 30–40 percent figure as a guideline. If the RTX 5070 were to cost $550, we wouldn't recommend spending more than $425 on the RTX 5060 Ti 16GB — and $400 would be better. Conversely, if the best price you can find on the RTX 5070 ends up being $700, then the 5060 Ti 16GB would be worth $500 to at most $540. (And you can replace the dollar signs with whatever monetary symbol you choose.)

Long-term, we expect both GPUs will trend toward about a 30% higher price for the RTX 5070 vs RTX 5060 Ti 16GB. Anything less makes the slower card very difficult to justify, even if it does technically offer more VRAM. So unless you specifically need a card with 16GB, perhaps to run a particular LLM, the RTX 5070 offers the better overall value right now, and will probably continue to do so in the future.

Speaking in an interview with Digital Foundry at CES 2025, Nvidia VP of Applied Deep Learning Research Bryan Catanzaro hinted that Nvidia may end up bringing DLSS 4 to Nvidia GPUs as old as the RTX 30 Series, which otherwise do not support Nvidia's own Frame Generation technologies. Instead, RTX 30 Series GPUs are currently forced to rely on games with an AMD FSR 3 Frame Generation solution — which isn't always usable alongside DLSS, though fortunately it usually is. We note that external solutions like Lossless Scaling, which also allows for Multi-Frame Generation like DLSS 4 will, though of course continue to work without AI acceleration.

In the full interview, the specific question asked of Catanzaro on Frame Gen for older Nvidia GPUs by Digital Foundry's Alex Battaglia was "Now that it (DLSS 4) is running entirely on Tensor cores, obviously it's a more intensive task as a result of that, to some degree... what's keeping it, then, from running on RTX 3000?"

Catanzaro replied, "I think this is primarily a question of optimization and engineering and then the ultimate user experience. We're launching this Frame Generation, the best Multi-Frame Generation technology with the 50 Series, and we'll be able to see what we squeeze out of older hardware in the future."

While this is hardly a conclusive statement, the logic leading to the question and the open-endedness of the answer does lead to some optimism. And, of course, Nvidia RTX 30 Series users can always opt for AMD FSR 3 Frame Generation wherever it's available, with Lossless Scaling's recent updates further improving its support for global Multi-Frame Generation.

Aside from this question, which could allude to some nice future boons for older hardware, Battaglia also asks several other questions in the full interview that dive deeper into the technical details of DLSS 4.

Highlights include Catanzaro claiming that DLSS 4 Frame Generation's frame pacing should be much more evenly paced than DLSS 3 Frame Generation, claiming a decrease in frame pacing variability by "a factor of 5 or 10." Basically just meaning that Frame Gen should now look far smoother. Though for now, Frame Generation still doesn't include V-Sync, and when asked, Catanzaro essentially said that the demand would need to be there for Nvidia to have the incentive to get it figured out.

A bracket for the unreleased Nvidia GeForce RTX 4090 Ti, which would have been one of the best graphics cards, has surfaced on the Chinese e-commerce platform Taobao. It is priced at 99 Chinese Yuan, or around $14. Hardware leaker Harukaze5719 shared the listing on X (formerly Twitter) with a link.

The I/O bracket shows GeForce RTX 4090 Ti, with a PG136F model number, which Nvidia has never released on any GeForce RTX 40-series GPU. There was much speculation as to why Nvidia never released the GeForce RTX 4090 Ti when we typically get an x090 Ti flagship with every generation of GeForce graphics cards.

Some claim that Nvidia should not have released a faster GPU due to the lack of competition. The current GeForce RTX 4090 sits alone on the performance throne, so releasing a quicker graphics card didn’t make sense. Others say that Nvidia is better off concentrating on professional-grade or AI GPUs, which have become a commodity in the AI boom.

However, the reason appears to be the U.S. ban on the RTX 4090 GPUs. MEGAsizeGPU, a reputable hardware leaker, claims that Nvidia reportedly canceled the GeForce RTX 4090 Ti due to the sanctions on the RTX 4090. China is one of the Nvidia’s largest GPU customers. With Washington’s move to stop the sales of Nvidia’s most powerful GPUs to the country, it could likely mean that the RTX 4090 Ti’s most prominent potential market is no longer legally accessible, making the graphics card commercially unviable.

While the GeForce RTX 4090 Ti probably existed, MEGAsizeGPU believes the specifications were never finalized. However, as mentioned in many leaks, the leaker thinks it would sport 24GB of memory. Meanwhile, others believe that Nvidia purportedly faced issues while optimizing the thermal aspect of the GeForce RTX 4090 Ti. There may be some truth to the rumors, as all the alleged photographs of the Ada Lovelace graphics card revealed a beefy triple-slot cooler.

This GeForce RTX 4090 Ti bracket will likely mean nothing for most users. It’s a collector’s item, like EVGA’s unreleased GeForce RTX 4090, for hardcore enthusiasts and hardware collectors. But at less than $15, it’s your chance to get a piece of graphics card history with an exciting story.

Driven by unprecedented demand for AI GPUs, Nvidia on Wednesday posted its all-time-record quarterly revenue of $30.040 billion. Yet its gross margins dropped a bit as the company prepped for its Blackwell product rollout in the fourth quarter. Surprisingly, even Nvidia's gaming revenue was up in the second quarter, which is not typical at all. 

Nvidia's revenue for the second quarter of its fiscal 2025 reached $30.040 billion, up 122% year-over-year and 15% quarter-over-quarter. The company's net income for the quarter totaled $18.642 billion, up a modest 10% sequentially but a whopping 174% compared to the same quarter a year ago. However, Nvidia's gross margin for the quarter was 75.1%, up 5% from Q2 FY2024 but down 3.3% from Q1 FY2025, as the company produced a low-yield Blackwell design in preparations for volume launch late this year.

Data Center

Data center business accounts for 87.4% of Nvidia's revenue. Last quarter, Nvidia sold some $26.272 billion worth of data center hardware (up 16% sequentially and up 154% year-over-year), which included compute hardware (primarily GPUs) worth $22.604 billion. Networking revenue reached $3.7 billion, a 114% increase compared to the previous year, fueled by InfiniBand and AI-related Ethernet revenue, including the Spectrum-X end-to-end Ethernet platform. On a sequential basis, networking revenue grew by 16%. 

The company increased sales of its AI and HPC GPUs by 162% compared to the same quarter a year ago, even though its sales to Chinese entities are now seriously curbed. 

During the quarter, the company began shipping its H200 GPU with 141 GB of HBM3E memory in volume, which helped to elevate sales. With yield issues plaguing the first version of Nvidia's Blackwell-based B100/B200 GPU, Nvidia's Hopper products will probably remain the company's working horses this fiscal year. Still, demand for these proven products is strong. 

"Hopper demand remains strong, and the anticipation for Blackwell is incredible,” said Jensen Huang, chief executive of Nvidia. "Nvidia achieved record revenues as global data centers are in full throttle to modernize the entire computing stack with accelerated computing and generative AI."

Graphics

On the gaming side, Nvidia's sales catapulted to $2.88 billion, an increase of 9% sequentially and 16% year-over-year. This is not in line with seasonality, as gaming graphics card sales are usually down in the second quarter, but Nvidia said that it sold quite some GPUs in July as part of the back-to-school season. In the high end, the company probably stole some market share from AMD.

Professional Visualization revenue increased to $454 million, 20% year over year and 6% sequentially. These gains were primarily driven by the ongoing growth of RTX GPU workstations built on the Ada Lovelace architecture.

Automotive, OEM, other

Nvidia's automotive revenue was up 5% sequentially and 37% year-over-year and totaled $346 million, thanks to sales of self-driving platforms and AI cockpit solutions. Nvidia's sales to OEMs reached $88 million, up 13% QoQ and 33% YoY, as PC makers finished their inventory corrections.

Q3 Guidance

The guidance for the third quarter is perhaps the weakest part of Nvidia's report. The company expects its sales to be around $32.5 billion ±2%, with GAAP and non-GAAP gross margins expected to be 74.4% and 75.0%, respectively. While the company's margins remain the highest in the industry, the modest revenue growth prediction worries Wall Street analysts. 

Then again, demand for Nvidia's Hopper and Blackwell products exceeds supply, so the humble guidance could reflect TSMC's ability to supply Nvidia with enough processors.

Nvidia is rumored to be working on a new RTX 3050 A laptop GPU powered by an AD106 (Ada Lovelace) die. It departs from the usual Ampere chip used by all other GPUs in the RTX 30-series lineup. The RTX 3050 A doesn’t have an official release date, nor did Nvidia announce it. However, the GPU is already included in Nvidia’s latest driver release and appears as the “NVIDIA GeForce RTX 3050 A Laptop GPU” in the PCI ID database.

The use of the AD106 die for the RTX 3050 A is an intriguing choice, as it has more transistors and CUDA cores than the GA107 die it replaces in the current RTX 3050s and the AD107 used in the RTX 4050 laptop. The AD106, currently used in the RTX 4060 Ti desktop and RTX 4070 laptop GPUs, has 22.9 billion transistors and 4,608 CUDA cores. On the other hand, the Ampere GA107 die only has 8.7 billion transistors and 2,560 CUDA cores, while the Ada Lovelace AD107 die has 18.9 billion transistors and 3,072 CUDA cores.

While we won’t complain about an entry-level laptop GPU getting a potentially faster GPU with additional features, it's not exactly clear what aspects of the chip remain enabled and it lands in something of an awkward spot. The RTX 3050 A laptop GPU could potentially rival the RTX 4060 in both laptop and desktop versions, which use the less powerful AD107 die, but that seems unlikely. Aside from the AD106 die that Nvidia plans to put into the RTX 3050 A, we have no other information about its specifications.

Given past behavior, it's a safe bet that Nvidia isn't planning on using a fully enabled AD106 chip in the RTX 3050 A. In fact, it's probably a dramatically cut-down version of the chip that wasn't able to meet the requirements of higher tier parts. AD106 normally goes into the desktop RTX 4060 Ti as well as the mobile RTX 4070 Laptop GPU. It features up to 36 SMs (Streaming Multiprocessors) in its maximum configuration, but inevitably there will be chips that have multiple faulty SMs. These can be disabled to create a lower tier part, and that's likely happening here.

What's not clear is whether the RTX 3050 A would still have the higher performance OFA (Optical Flow Accelerator) that used for DLSS 3, and if it will support that. Probably not is our guess, but we'll have to wait and see how this plays out. Note that the standard RTX 3050 Laptop GPU also only comes with either 4GB of memory on a 128-bit interface, or 6GB on a 96-bit interface, so Nvidia could go either direction with the 3050 A.

We’re unsure whether Nvidia will even bring this GPU to market, but adding new SKUs this late in its lifespan isn’t unheard of. For example, we received the entry-level Nvidia GTX 1630 in late 2022 — two years after the RTX 30-series parts started shipping and over three years after the GTX 1650 using the same TU117 die launched. What we do know is that, in general, Nvidia tends to use similar names for similar levels of performance (within about 5%), so the RTX 3050 A Laptop GPU, if it does show up in systems at some point, likely won't be too much faster than the existing Ampere-based parts.

Whatever else might be disabled, AD106 uses TSMC's superior 4N node, rather than the Samsung 8N node used by Ampere, with nearly twice the power efficiency — meaning 3050 A could improve battery life if nothing else. It will be interesting to see how the RTX 3050 A performs compared to the existing RTX 3050 and RTX 4050 laptops, if or when it ships.

There are a lot of portable notebooks suitable for gaming, but when it comes to truly high-performance machines that can rival desktops, they tend to be pretty large. Acer this week introduced its Predator Helios Neo 14 that offers high-end components and a great display in a rather compact 14-inch chassis. In addition, Acer launched its new AMD and Nvidia-based Nitro 14 laptop that offers performance at a more affordable price.

Acer Predator Helios Neo 14

Acer's Predator Helios Neo 14 comes equipped with a 14.5-inch IPS display with an up to 3072x1920 resolution and a variable refresh rate of up to 165 Hz with Nvidia G-Sync on top. The laptop weighs 1.9 kilograms (4.19 pounds) and is up to 20.9 mm (0.82 inches) thick. 

The Predator Helios Neo 14 packs up to Intel's Core Ultra 9 185H (16 cores, 6P + 8E + 2LP at up to 5.10 GHz) paired with up to 32 GB of LPDDR5X memory,  up to 1 TB PCIe Gen4 SSD as well as up to Nvidia's GeForce RTX 4070 8GB Laptop GPU. To ensure consistent performance of the system under high loads in games or demanding applications, the laptop is outfitted with an advanced cooling system with heat pipes featuring two 5th Gen AeroBlade 3D fans and Vortex Flow technology as well as liquid metal thermal grease on the CPU. 

As for connectivity, Acer's compact gaming laptop has everything that one comes to expect from a powerful 2024 gaming notebook, including Intel's Killer 1675i Wi-Fi 6E + Bluetooth adapter, a Thunderbolt 4 port, two USB Type-A connectors, two display outputs (one DisplayPort using a USB-C connector and one HDMI), a microSD card reader, a 1080p webcam, and a 3.5 mm connector for headsets.  

The machine is equipped with a 76 Whr battery, and Acer bundles a 230W proprietary AC adapter for fast charging and gaming.

The Predator Helios Neo 14 (PHN14-51) is set to be released in North America in May with a starting price of $1,799.99, and in EMEA in June with a starting price of €1,999. The entry-level configuration packs Intel's Core 5 Ultra 125H (14 cores, 4P + 8E + 2LP cores at up to 4.50 GHz) CPU and Nvidia's GeForce RTX 4050 6 GB GPU, so it remains to be seen how much top-end versions will cost.

Acer Nitro 14

In addition to launching its top-of-the-range 14.5-inch Predator Helios Neo 14 laptop, Acer also launched its lower-cost 14.5-inch Nitro 14 machine that promises to offer performance gaming at a considerably lower cost.

Acer's Nitro 14 is based on an up to AMD's six-core Ryzen 7 8845HS processor with Nvidia's GeForce RTX 4060 8 GB Laptop GPU, up to 32 GB of LPDDR5X memory, and an up to 1 TB PCIe Gen 4 SSD. On the connectivity front, the Nitro 14 has a Wi-Fi 6E + Bluetooth adapter, a USB 4 port, two USB Type-A connectors, two display outputs (one DisplayPort using USB-C, one HDMI), a microSD card reader, and an audio jack.

The machine comes with a 14.5-inch display featuring an up to 2560x1600 resolution and an up to 120 Hz variable refresh rate with Nvidia's G-Sync. Compared to the Predator Helios Neo 14, the Nitro 14 is both heavier (2.02 kilograms, 4.45 pounds) and thicker (22.84 mm, 0.9 inches), but it is much more affordable too.  

The Acer Nitro 14 (AN14-41) is set to be available in North America in May with a starting price of $1,299.99, and in EMEA in June with a starting price of €1,199.

16-Inch Nitro Laptops Get a Refresh

In addition to launching its all-new Predator Helios Neo 14 and the Nitro 14, Acer also refreshed its 16-inch Nitro 16-series machines with new CPUs. 

The Nitro 16 (AN16-73) has been updated with up to an Intel Core i7-14700HX processor and up to an Nvidia GeForce RTX 4060 Laptop GPU. The machine will be available in North America in May, starting at $1,399.99, and in EMEA in June, starting at €1,599 for an entry-level configuration. 

The AMD-based Nitro 16 (AN16-43) is now equipped with an up to an AMD Ryzen 9 8945HS processor and up to an Nvidia GeForce RTX 4060 Laptop GPU. The machine also features a 16-inch WQXGA (2560x1600) display with 500 nits of brightness and a 165 Hz refresh rate. This unit will be available in North America in May, starting at $1,399.99.

Nvidia has launched two new entry-level GPUs for professionals. The new RTX 500 and 1000 Ada Generation laptop GPUs expand the mobile RTX Ada Generation line-up to seven, and these particular models are particularly suited to thin and light designs, according to Nvidia. The potential 154 and 193 TOPS delivered by the RTX 500 and 1000, respectively, is claimed to be enough for content creators, researchers, or engineers on the go. Even the humble RTX 500 delivers “up to 14x the generative AI performance,” compared to CPU-only configurations, according to Nvidia's in-house testing.

With the two new RTX Ada Generation GPUs for laptops, the family expands to seven SKUs. However, Nvidia is also currently maintaining previous-generation Ampere laptop chips with similar naming. Check out the table below for the full range, and the respective product specifications.

Above you can see the specs and a couple of performance metrics for the workstation laptop GPUs. There have been some significant uplifts in performance delivered by the new Ada Generation GPUs, so you wouldn’t want to mix up an Nvidia RTX 500 Ada Generation (154 TOPS) with its Nvidia RTX A500 predecessor (56 TOPS), for example. The GPU names are very similar if you don't know what to look for.

The new RTX 500 and 1000 Ada Generation laptop GPUs share some key similarities. They are both based on Nvidia’s AD107 GPU, for example. Moreover, their CUDA core counts are booth in the 2000s, at 2,048 and 2,560, respectively. Memory subsystems are also quite anaemic with the RTX 500 Ada Generation sporting 4GB of GDDR6 on a 64-bit bus, and the RTX 1000 Ada Generation only raising this to 6GB of GDDR6 on a 96-bit bus. However, such low specs help keep the power budget down, with the lowest-end chip only pulling 35-60W. The RTX 1000 Ada Generation ramps up from 35-140W with up to 15W extra Dynamic Boost. Overall, this hardware clearly targets the stated thin and light workstation segment.

In its press release, Nvidia also sought to remind potentially interested parties that increased TOPS isn’t the be-all and end-all of the new Ada Generation chips. Like the existing members of the product family these two little laptop GPUs also offer the benefits of: third-gen RT cores, fourth-gen Tensor cores, Ada Generation CUDA cores, DLSS3, the eighth-gen NVENC with AV1 support, and beat iGPUs thanks to their dedicated GPU memory.

Portable workstations packing RTX 500 and 1000 Ada Generation laptop GPUs from partners such as Dell, HP, Lenovo, and MSI are expected to become available this spring.

Alongside pre-launch retailer listings, VideoCardz has unearthed a low-res render of a purported RTX 2000 Ada desktop workstation GPU. The source reckons the release date for this entry-level Ada workstation card with 8GB of VRAM is sometime during the coming week. 

This new RTX 2000 workstation GPU seems to be yet another AD104-based GPU, with 8GB of (likely ECC) GDDR6 VRAM and PCIe Gen 4 x16 support. It also appears to lack any external power connector, so we should expect a TDP of no greater than 75 Watts (the usual maximum power draw of a PCIe slot).

Two model names listed in the VideoCardz report are the Nvidia RTX 2000 ADA and the Nvidia RTX2000E ADA, with estimated price points of $600-650 and $500-630, respectively. What the "E" suffix signified is currently unconfirmed, but "E" may refer to "Enterprise". It is the cheaper of the two models, though, so perhaps that model is only available when purchasing in bulk.

Since the AD104 is the GPU used in the consumer GeForce RTX 4070 and RTX 4080, this workstation card should in theory be quite powerful for its (potentially SFF) form factor and low wattage. Reducing wattage to this extent will greatly reduce performance, but AD104 provides a strong foundation. Additionally, all of Nvidia's workstation (previously dubbed Quadro) GPUs are sold with blower-style coolers, making them more favorable for multi-GPU stacking in servers and other workstation scenarios.

At the time of writing, an entry-level '2000' model is conspicuous by its absence from Nvidia's listings of Ada architecture RTX GPUs for workstations. The purported Ada Generation RTX 2000 GPU specs line up with a long-due refresh of Nvidia's entry-level workstation video card offerings.

With any luck, this rounding out of the Ada Lovelace workstation offerings may also push Nvidia toward considering new branding for its next generation of workstation GPUs. The naming scheme of workstation Nvidia GPUs has gotten increasingly contrived gen-over-gen since the change, particularly since mainstream RTX GPUs use a nearly identical naming scheme, except focusing on a fixed RTX XX-Y0 range instead of RTX X-000 for each model number.

Needing to add 'Ada Generation' to every product listing made for a current-generation Nvidia workstation GPU seems a little excessive for a system that used to use Quadro branding and some numbers more effectively. Still, we'd be veering off topic if we lingered on this any longer. It is worth mentioning now that Nvidia's Ada Lovelace lineup seems complete, though.

A premium GeForce RTX 4080 Super graphics card equipped with a 280mm AiO liquid cooler is being teased in China (h/t WCCFTech). Newcomer Zephyr has collaborated with the well-respected Valkyrie (VK) for this distinctive, unique silver and black design, which boasts several RGB LED lighting zones. The resulting Zephyr X Valkyrie RTX 4080 Super is demoed in the linked video and appears to be a cool and pleasingly quiet performer. But this is just a teaser, of course — we're still waiting for official tech specs, pricing, and availability.

When the Zephyr X Valkyrie RTX 4080 Super is released it will be the first liquid-cooled graphics card from Zephyr to come to market. The design is based on one of the best graphics cards currently available and is said to be the result of six months of close collaboration between Zephyr and Valkyrie.

The predominantly silver GPU shroud features a series of design flourishes: windows with LED logos, venting, and multi-faceted bevels — but it isn’t too busy-looking.

Zephyr and Valkyrie keep the top of the design clean by putting both the 12VHPWR and USB connectors on the end of the card. There is also supposed to be a full-length backplate, so the custom PCB is completely encased.

The radiator connects to the GPU via colored braided tubes, and looks like it uses similar materials to match the rest of the white and silver demo system PC build. The 280mm radiator makes use of twin 140mm fans with silvery centers, surrounded by angular shards of LED lighting.

Several other components in this PC are from the experts at Valkyrie, and everything sits together in a pleasantly unified theme.

In the video, Zephyr thanks Valkyrie for its cooling technology, which it says enables the RTX 40080 Super to “run at extremely low temperatures while exerting powerful performance, resulting in an excellent, quiet experience.”

Backing up that statement, there's some video footage showing 3DMark benchmarking, FurMark purring along, and various games running on demanding settings. We can see from the video that the GPU temperature never exceeds 52 degrees Celsius. Moreover, the fan speed doesn't require pushing past 1,350 rpm to maintain this max temperature, and the system noise measures around 40dB (similar to a quiet library). As always, take these first-party test results and measurements with a pinch of salt.

If you're interested in the Zephyr X Valkyrie RTX 4080 Super, the video says buyers should expect pre-orders to open shortly after Chinese New Year. We expect we'll see more collaborative work between these hardware makers, but whether any Zephyr designs become available in the west remains to be seen.

Nvidia's GeForce RTX 4080 Super graphics cards has two significant differences compared to its non-Super predecessor: higher performance and a $200 lower recommended price. That's almost an unexpected combination in the current market for standalone desktops graphics, though it's also a tacit admission that the 4080 was originally priced too high. Apparently, there is a third difference as well: a simplified design of Nvidia's own GeForce RTX 4080 Super Founders Edition board.

Nvidia's GeForce RTX 4080 Founders Edition graphics card has a voltage regulating module featuring 13 power phases for its AD103 graphics processing unit, with three power phases for its 16GB of GDDR6X memory. By contrast, the new GeForce RTX 4080 Super Founders Edition board has an 11-phase VRM for the GPU and a 2-phase power supply circuitry for the memory, as noticed by reviewer Geekerwan (via @9550pro).

The slightly simplified printed circuit board (PCB) design does not really come as a surprise. On the one hand, by now Nvidia and TSMC have probably reduced performance variability of the AD103 graphics processor to a minimum and so Nvidia doesn't need to ensure the best possible power supply for the GPU — at least with 10,240 CUDA cores operating at up to 2,550 MHz — as virtually all chips will perform as advertised even with an 11+2 phase VRM. 

Meanwhile, simplifying the GeForce RTX 4080 Super Founders Edition makes the product slightly cheaper to produce, so Nvidia can sell one of the best graphics cards on the market for $999 instead of $1,199 and retain its profit margin. The slight design optimization should not tangibly affect reliability of the product.

What remains to be seen is how well Nvidia's GeForce RTX 4080 Super Founders Edition graphics cards overclock. Advanced power supply circuitry tends to have a larger effect on overclockability, so the reduction in the number of phases may somewhat lower ability of the card to work at clocks beyond its official specifications — not that we noticed any issues in our review, where the average core clocks on the 4080 Super were quite high.

Nvidia also has plenty of partners offering custom GeForce RTX 4080 Super graphics cards that not only come factory overclocked, but also feature enhanced PCB designs and massive cooling systems to ensure additional overclocking potential. We'll be looking at some of these in the coming days.

Avid readers of Tom's Hardware already know that graphics processing units (GPUs) designed for laptops tend to end up on desktop graphics cards made by little-known Chinese brands. Well, this has happened again with Nvidia's GeForce RTX 4080 Laptop and GeForce RTX 4090 Laptop GPUs that were installed on larger printed circuit boards (PCBs) and sold as desktop graphics cards, reports VideoCardz.

For obvious reasons, the so-called GeForce RTX 4080M (listed at $420 - $560) and GeForce RTX 4090M are cheaper than some of Nvidia's best graphics cards from renowned brands partnering with Nvidia. However, apart from build quality concerns (which is something one should consider with little-known makers of graphics cards), there are peculiarities to consider before buying one of these mobile-originating GPU-based products. Perhaps the main one here is performance: being mobile GPUs, they are cut down in terms of the number of active CUDA cores. 

Based on the VideoCardz report, the so-called GeForce RTX 4080M (GeForce RTX 4080 Laptop GPU) is said to perform somewhere between the GeForce RTX 4070 and GeForce 4070 Ti desktop models. Meanwhile, it has a lower power limit of 285W, which is still much higher than the mobile form factor's peak of 175W, and less memory than the real GeForce RTX 4080. As for the so-called GeForce RTX 4090M, it may be at least 40% slower than the real GeForce RTX 4090 based on the number of CUDA cores alone. 

A short brief of specifications of the so-called GeForce RTX 4080M and GeForce RTX 4090M graphics cards looks as follows:

Of course, these graphics cards are relatively cheap, but there are downsides to buying these GPUs. Firstly, they have limited warranty support (since you essentially buy from a no-name brand). Secondly, Nvidia has different drivers for desktop and laptop GPUs, and at some point, this may become a burden.  

In general, while these mobile-turned-desktop GPUs can be tempting due to their lower price, it is important to be aware of their generally lower performance compared to legit desktop parts, as well as other risks before making a purchase.

Although small form-factor desktops are rather popular these days, makers of graphics cards, for some reason, refrain from making low-profile add-in-boards for gamers. Regarding the GeForce RTX 40-series, only Gigabyte introduced a low-profile GeForce RTX 4060 graphics board last August. Now Asus is catching up with its GeForce RTX 4060 LP BRK 8GB GDDR6 product that promises to pack a punch in a very compact package, making it one of the best graphics cards for SFF desktops.

The Asus GeForce RTX 4060 LP BRK 8 GB GDDR6 graphics card carries Nvidia's AD107 graphics processor featuring 3072 CUDA cores and 8 GB of GDDR6 memory connected to the GPU using a 128-bit interface. The board's graphics chip operates at up to 2490 MHz in OC mode, which is just 30 MHz higher than the frequency recommended by Nvidia. The graphics card boasts 15 FP32 TFLOPS of compute performance and 272 GB/s memory bandwidth. The compute performance of the board is 25% higher compared to the compute performance of the Xbox Series X GPU, but the console boasts 560 GB/s of peak memory bandwidth.

Interestingly, Asus positions its GeForce RTX 4060 LP BRK 8 GB GDDR6 as a 'low-profile, compact graphics card that enables multi-monitor productivity,' which implies that the add-in-board is generally aimed at users who need four display outputs (two DisplayPorts and two HDMI) and more or less decent performance in a compact form-factor. However, the board can be used for an SFF gaming PC or a home theater PC (HTPC).

Asus equipped its GeForce RTX 4060 LP BRK 8 GB GDDR6 graphics card with a dual-slot triple-fan cooling system with several heat pipes to ensure higher efficiency. Given that the board is rated for up to 115W total graphics power, such a cooling system should be enough to ensure stable and consistent operation. Meanwhile, the board requires a six-pin auxiliary PCIe power connector, which is easy to find inside a DIY desktop but may be absent in machines from a large OEM.

Asus did not announce the pricing of its upcoming products, so we can only make guesses about the GeForce RTX 4060 LP BRK 8 GB GDDR6 in retail. However, Gigabyte sells its Gigabyte GeForce RTX 4060 OC Low Profile 8G for $325, which is $25 higher than Nvidia's MSRP for its GeForce RTX 4060.

The preliminary specifications of Nvidia's GeForce RTX 4070 Super, poised to be among the best graphics cards, claim that the GPU could have 48MB of L2 cache. To our great surprise, the launch materials that Nvidia supplied us mentioned that the GeForce RTX 4070 Super was equipped with a GPU with 36MB of L2. This was a typo later corrected by Nvidia, reports Wccftech.

Nvidia's GeForce RTX 4070 Super graphics card is based on the AD104 graphics processor with 7,168 CUDA cores enabled and a 192-bit memory interface. This is not the full AD104 GPU — used on the GeForce RTX 4070 Ti — with 7,680 CUDA cores and a 192-bit memory interface. However, the AD104 used for the GeForce RTX 4070 Super comes with a full 48MB L2 cache and, therefore, enjoys all the memory bandwidth benefits it can provide.

We can only wonder why Nvidia made a typo in its launch materials. Still, the company's website description of its GeForce RTX 4070 Super notes that the product has 48 MB of L2 cache and a memory subsystem featuring 504 GB/s of peak bandwidth.

The compute performance of Nvidia’s GeForce RTX 4070 Super is already very close to that of the GeForce RTX 4070 Ti: 35.5 FP32 TFLOPS versus 40.09 FP32 TFLOPS. With the same memory subsystem, the new GeForce RTX 4070 Super board promises to offer performance that is very close to that of the GeForce RTX 4070 Ti. The difference between the two products can be shrunk even further with some overclocking.

Remember that Nvidia and TSMC refine process technologies to improve yield and reduce performance variability, sometimes leading to higher overclocking potential for GPUs. Therefore, the overclocking potential of the AD104 GPUs produced today is probably higher than that of AD104 graphics processors produced in late 2022 or early 2023.

Therefore, perhaps the most exciting part will be to learn how custom GeForce RTX 4070 Super graphics cards from Nvidia’s AIB models will behave. Some of those boards will have enhanced TGP (I.e., higher than 220W set by Nvidia), and they will likely also enjoy GPUs with higher overclocking potential.

Zotac will not charge extra for custom versions of its GeForce RTX 40-series Super graphics cards that come with enhanced coolers and will sell them at Nvidia's recommended retail prices, the company announced on its website (as noticed by @momomo_us). Meanwhile, the company's factory-overclocked graphics cards will be sold at a premium given their higher performance. 

Zotac will offer Trinity Black Edition and Twin Edge versions of its GeForce RTX 4070 Super, GeForce RTX 4070 Ti Super, and GeForce RTX 4080 Super graphics cards at Nvidia-set MSRPs: $599, $799, and $999, respectively. Meanwhile, factory-overclocked Trinity Black Edition OC, Trinity White Edition OC, Twin Edge OC, and AMP-badged graphics boards will be sold at a premium, the company announced. All of these boards — pre-overclocked or not — are contenders to join the ranks of the best graphics cards available today. 

Although AMD and Nvidia offer reference designs for almost all of their graphics cards to speed up time-to-market, add-in-boards (AIBs) manufacturers, such as Zotac, tend to develop custom cards for two reasons: to differentiate themselves from their rivals with better looks and higher performance as well as to cut down costs. 

Given the fact that factory-overclocked versions of Trinity Black Edition and Twin Edge will use the same printed circuit boards and cooling systems as their counterparts that function at Nvidia-recommended frequencies, it does not look like regular versions were designed with the sole purpose of cutting down costs. As a result, regular Trinity Black Edition and Twin Edge models will likely feature decent overclocking potential (courtesy of a refined PCB and an enhanced cooling system), which makes them a good deal, assuming that they will be available at MSRP. 

Meanwhile, those who want to have enhanced performance right out of box and some further overclocking potential will be able to buy Zotac's GeForce RTX 4070 Ti Super AMP Holo 16GB GDDR6X or GeForce RTX 4080 Super AMP Extreme Airo 16GB GDDR6X that promise to feature significantly improved clocks and a refined coolers.

Gigabyte is prepping four GeForce RTX 4070 graphics card designs, with only the WindForce and Eagle models shipping with standard 8-pin auxiliary PCIe power connectors, while the Gaming and Aero models will have 12VHPWR 16-pin connectors. All four designs will have three fans, based on the leaked images. 

One of the things that strikes the eye with the pictures of the upcoming GeForce RTX 4070 graphics cards from Gigabyte, Palit, MSI, and Zotac is that most of them are triple-wide with a massive cooling system, just as my colleague Mark reported yesterday about the hoard of MSI's GeForce RTX 4070 AIBs.

While it is not particularly surprising that many of GeForce RTX 4070-based graphics cards will be huge, considering the overclocking potential of the GPU that powers them, it is reasonable to expect more compact AIBs too. This is perhaps where MSI's GeForce RTX 4070 Ventus 2X comes into play. The board features a dual-fan two-wide cooling system and will likely fit into most of PCs used by gamers. Of course, different GeForce RTX boards feature different power connections, some keep using well-proven eight-pin plugs, while others employ the all-new 12VHPWR.

Apparently, budget-friendly options from Zotac and Palit will also have dual-fan designs and 12GB of GDDR6X memory, yet take the information with a grain of salt as this comes from unofficial sources.

The Nvidia RTX 4070 Super, RTX 4070 Ti Super, and RTX 4080 Super officially broke cover today, months after the first leaks and rumors started circulating. To the surprise of virtually no one, the new GPUs offer a decent bump in performance over the existing non-Super counterparts, and will compete with the best graphics cards starting later this month.

Nvidia will be launch the cards in reverse order of performance, so the 4070 Super comes first with the 4080 Super as the caboose — and not coincidentally, that’s also the roughly the order of excitement we have for the three new models as well.

Nvidia appears to be following the pattern of its RTX 20-series GPUs from 2018/2019 with the RTX 40-series, with a few minor tweaks to the formula. At launch, just like the 20-series, most of the RTX 40-series GPUs were criticized for offering a relatively small bump in performance at higher price points than their predecessors. Another major complaint was with VRAM capacities and memory interface widths. That at least is one aspect that will change slightly with the RTX 40-series Super models.

Let’s go ahead and hit the speeds and feeds. Nvidia hasn’t provided exact specifications for the three new parts, but based off the performance figures and what we do know, we’ve filled in a few blanks — these approximations are indicated by the tilde (~) symbol before the clock speeds and core counts. We’re relatively certain on the specs, though it’s possible we could see a +/- 2 SM change on some of the parts.

Nvidia GeForce RTX 4070 Super

Let’s start with the RTX 4070 Super, which will use the AD104 GPU once more. The new GPU offers up 36 of FP32 shader teraflops, and 568 FP8 (with sparsity) AI teraflops. Those are important qualifications, as currently there are a lot of AI workloads that don’t leverage the FP8 format and/or sparsity, and turning off each of those cuts the theoretical performance in half — meaning for generic FP16 workloads running on the tensor cores, it has 142 teraflops of compute. We expect the reference 4070 Super to come with 56 Streaming Multiprocessors (SMs), giving it 7168 CUDA cores.

The memory interface will remain 192-bit, with the same 21 Gbps GDDR6X chips as the RTX 4070 and 4070 Super. That gives it the same 504 GB/s of bandwidth as the existing parts, with 22% more compute performance than the vanilla RTX 4070 and 11% less compute performance than the RTX 4070 Ti. It will come with the full 48MB of L2 cache as well (Nvidia's slide incorrectly states 36MB, we were told.)

Keep in mind that the official boost clock of around 2.48 GHz will likely be very conservative — our testing of previous RTX 40-series GPUs show most will run at 2.7–2.8 GHz — so it’s possible and even likely that the 4070 Super will land a lot closer to the 4070 Ti than the raw specs indicate. TGP gets a slight bump to 220W as well, which again should give it some legroom over the base 4070 but keeps it far behind the 4070 Ti’s 285W TGP.

With the bump in performance, Nvidia says the RTX 4070 Super will slightly outperform the previous generation RTX 3090, even if you don’t factor in DLSS 3 frame generation. We suspect it will largely depend on the games you play, or the games we use for testing, but the 4070 Ti was slightly faster than the 3090 so the numbers look about right.

Nvidia will be making an RTX 4070 Super Founders Edition card, which appears virtually identical in design to the RTX 4070 Founders Edition, other than a change in the color and casing of the shroud. The new 4070 Super comes dressed in stealthy black, and there’s no indication it will have any RGB lighting. That again echoes what we saw with the RTX 4070 and 4060 Ti Founders Edition cards.

The RTX 4070 Super will launch on January 17, with an MSRP of $599, taking over the price point of the RTX 4070. To make room for the newcomer, the RTX 4070 MSRP will drop to $549, which doesn’t seem like quite enough, and if you check current retail prices, you’ll see RTX 4070 cards already starting at $549 and sometimes less. We expect performance to land relatively close to the RTX 4070 Ti, which means you get a substantial improvement in bang for the buck, as the 4070 Ti MSRP currently sits at $799.

Nvidia GeForce RTX 4070 Ti Super

Moving on to the RTX 4070 Ti Super, things get a bit more interesting. Nvidia will use AD103 for the new GPU, opening the door for a 256-bit memory interface — a 33% improvement over the vanilla 4070 Ti. Given what we just discussed with the RTX 4070 Super, it’s no surprise that Nvidia indicates it will phase out the existing RTX 4070 Ti in favor of the new model, which will inherit the $799 MSRP price point. Where will fully enabled AD104 chips end up if the 4070 Ti goes away? Simple: They’re used in the mobile RTX 4080 as well as certain professional cards like the RTX 4000 Ada Generation.

Exact specifications have not been confirmed by Nvidia, but we do anticipate the 4070 Ti Super will have 66 SMs and 8448 CUDA cores, a 10% increase over the 4070 Ti. If we’re correct, that means the new GPU will have the same 2.61 GHz boost clock as the 4070 Ti, giving it 44 teraflops of FP32 shader performance and 706 teraflops of FP8 throughput — and the same 706 teraops of INT8 performance, if you’re wondering. TGP remains the same as the existing 4070 Ti at 285W.

The big change will be the memory bandwidth and capacity. Now Nvidia will have a 16GB card with a 256-bit interface and 672 GB/s of bandwidth (still 21 Gbps GDDR6X). That should make the 4070 Ti Super relatively competitive with AMD’s RX 7900 XT, and clearly ahead of the RX 7800 XT. In fact, the RTX 4070 Ti Super could land very close to the RX 7900 XTX in overall performance based on our existing test results — with the usual mix of superior ray tracing and AI performance, and slightly lower rasterization performance.

Nvidia does equip the 4070 Ti Super with 48MB of L2 cache — 6MB per 2GB memory channel, instead of the full 8MB per channel. This is clearly one small way to try and differentiate the 4070 Ti Super form the 4080/4080 Super. What we've seen with previous testing of RTX 40-series cards is that you get diminishing returns with larger L2 caches, so overall this likely only makes a few percent difference between the various AD103-based cards.

You can check our GPU benchmarks hierarchy for the full rundown, but the RTX 4080 already matches the RTX 7900 XTX in rasterization performance (give or take a couple percent), and leads by up to 35% in ray tracing performance, with a 10–15 percent lead overall. 

The RTX 4070 Ti Super looks like it will be about 10% slower than the RTX 4080 — or around 15% faster than the RTX 4070 Ti — which puts it right around the level of AMD’s top GPU, at a lower price (but with less VRAM). And Nvidia still has some extra features like DLSS support and its overall software and API ecosystem.

The RTX 4070 Ti Super is slated to launch one week after the RTX 4070 Super, on January 24. Unlike the other two Super cards, but just like the prior RTX 4070 Ti, there will only be third-party cards for the RTX 4070 Ti Super — so there will be no RTX 4070 Ti Super Founders Edition, probably because that’s just a few too many descriptors tacked onto the name.

Nvidia GeForce RTX 4080 Super

Wrapping things up, the RTX 4080 Super in many ways feels like the most disappointing refresh of the new cards. The RTX 4080 already used a nearly maxed out AD103 GPU with 76 SMs, and RTX 4080 Super will go whole hog and offer the full 80 SMs and 10240 CUDA cores. That’s only a 5% increase in core counts, if you’re keeping track, and AD103 means Nvidia is still stuck with a 256-bit memory interface and 16GB of VRAM — so my hopes for an RTX 4080 Super 20GB are dead and buried. Sigh.

Nvidia will give the RTX 4080 Super a slight bump in GDDR6X clocks, going from 22.4 Gbps on the vanilla 4080 to 23 Gbps on the 4080 Super. That’s also a pretty paltry 2.7% boost to bandwidth, which will hardly make any noticeable difference in most workloads. In fact, we suspect most third-party factory overclocked RTX 4080 cards will come very close to matching the RTX 4080 Super. The new GPU will also have the same 320W TGP as its predecessor.

The good news is that Nvidia is giving the RTX 4080 Super a much-needed price cut, and it will have a $999 MSRP, $200 less than the existing RTX 4080. As you can guess, that means the RTX 4080 will be phased out in favor of the new cards, though we could see a few short-term deals to clear out any remaining inventory. Or put another way: Don't pay more than about $899 for an RTX 4080 at this point.

Of course, no one is going to be looking at upgrading from an RTX 4080 to the RTX 4080 Super, so Nvidia provided performance comparisons with the existing RTX 3080 Ti and RTX 2080 Super. If you have the latter and skip a generation to upgrade to the 4080 Super, you should see more than a doubling of performance… along with a 43% increase in launch pricing, if you’re keeping score. You also get double the VRAM, and in AI workloads Nvidia says the 4080 Super outperforms the 3080 Ti by 50–70 percent.

The RTX 4080 Super will launch on January 31, one week after the RTX 4070 Ti Super.

Once the three new cards launch, by the end of January, here’s what Nvidia’s RTX 40-series lineup will look like. Again, note that the RTX 4080 and RTX 4070 Ti vanilla cards will be phased out in favor of newer models. There’s no Super upgrade for the 4060 Ti and 4060, though in our view we could certainly see the RTX 4070 prices drop closer to the $499 mark, and that would be just as good as a hypothetical RTX 4060 Ti Super.

Besides the new RTX 40-series Super GPUs, Nvidia has some other announcements regarding GeForce Now, new RTX-enabled games, new AI technologies, and livestreaming updates. We'll cover those separately in a different article, though you can find the relevant slides in the full presentation deck that we’ve included below.

Zotac is prepping three GeForce RTX 4070 Super graphics cards, according to a report by VideoCardz. The boards will use rather sophisticated cooling systems as well as a 12VHPWR power connector, which is a bit different from the company's original GeForce RTX 4070 lineup, which included boards with an eight-pin power plug.

Zotac will have Trinity Black and Twin Edge versions of GeForce RTX 4070 Super graphics cards. The GeForce RTX 4070 Super Trinity Black is a more premium product with a large three-wide triple-fan cooling system and one 12VHPWR power connector. By contrast, the GeForce RTX 4070 Super Twin Edge will be aimed at more price-conscious buyers and will be compact, with a two-wide dual-fan cooling system featuring multiple heat pipes.

While Zotac's original (non-Super) Twin Edge GeForce RTX 4070 used eight-pin auxiliary PCIe power connectors, the Twin Edge GeForce RTX 4070 will exclusively use 12VHPWR connectors — which may affect its compatibility with older systems. 

Nvidia will reportedly introduce its GeForce RTX 4080 Super 16GB, GeForce RTX 4070 Ti Super 16GB, and GeForce RTX 4070 Super 12GB at an event on Jan. 8.

The GeForce RTX 4070 Super 12GB is expected to be based on the AD104-350/AD103-175 with 7168 CUDA cores and a 192-bit memory interface. The product will be slightly more power-hungry than the non-Super version (225W vs 200W) and will cost $599 - $649, according to rumors. 

Alleged Nvidia RTX 40-Series Super Specifications

*Specifications are unconfirmed.

With 7168 CUDA cores active, Nvidia's GeForce RTX 4070 Super has the potential to become one of the best graphics cards around — especially at its price point. Zotac, one of the closest partners of Nvidia, does not plan to offer heavily factory-overclocked GeForce RTX 4070 Super versions (belonging to the AMP family of products), at least according to the VideoCardz report.

A day after we reported that Gigabyte is prepping a bunch of GeForce RTX 40-series Super graphics cards, the company's gaming division Aorus published a teaser depicting a new graphics board at CES. Given that AMD is not expected to release any new high-end Radeons at the upcoming tradeshow, it is more than likely that Aorus Turkey will essentially confirm Nvidia's RTX 40 Super launch.

The graphics card Aorus Turkey showcases is akin to other Gigabyte's Aorus Master WindoForce-branded high-end offerings: at least three slots wide. It has three 'bionic shark' fans. Meanwhile, there is an RGB ring around each fan. It is, of course, impossible to tell what GPU the board on the picture carries and whether the cooling system of the alleged Aorus GeForce RTX 40-series Super graphics card is different from 'non-Super' offerings. Still, at least the upcoming products are similar to existing boards design-wise.

Nvidia is expected to introduce its GeForce RTX 4080 Super, GeForce RTX 4070 Ti Super, and GeForce RTX 4070 Super at a special event on January 8, a day before CES kicks off. Since Nvidia takes the wraps of its 'Super' products on January 8, it is reasonable to expect its partners to demonstrate their versions of these devices starting the next day at the trade show.

Alleged Nvidia RTX 40-Series Super Specifications

*Specifications are unconfirmed.

Nvidia's GeForce RTX 4080 Super, GeForce RTX 4070 Ti Super, and GeForce RTX 4070 Super are reportedly based on the AD103 and AD104 graphics processors featuring the Ada Lovelace architecture.

While the GeForce RTX 4080 Super is expected to offer a moderate performance bump compared to the non-Super model, the GeForce RTX 4070 Ti Super and GeForce RTX 4070 Super are projected to be noticeably faster than their non-Super counterparts, at least based on leaked specifications.

Numerous listings of Nvidia's pending GeForce RTX 40-series Super graphics cards have appeared in the Eurasian Economic Commission (EEC) database. With all the chatter regarding Nvidia's refresh of its desktop lineup as well as the current listings, we can be almost sure that the company is indeed gearing up to introduce its GeForce RTX 4080 Super, GeForce RTX 4070 Ti Super, and GeForce RTX 4070 Super at an event on January 8. Apparently, Nvidia is also prepping the GeForce RTX 3050 6GB.

@Harukaze5719, a well-known hardware detective, has noticed that Gigabyte (or someone close to the company) has registered a variety of the company's GeForce RTX 40-series Super graphics cards with the Eurasian Economic Union (EEU) customs, confirming their memory configurations. The products in question are the GeForce RTX 4080 Super 16 GB, GeForce RTX 4070 Ti Super 16 GB, GeForce RTX 4070 Super 12 GB, and GeForce RTX 3050 6 GB. These GeForce RTX 40 Super-series graphics boards can be among the best graphics cards if the pricing is right. 

Meanwhile, VideoCardz has noticed that a Swiss retailer has listed multiple GeForce RTX 40-series Super graphics cards with similar memory configurations, seemingly corroborating the EEC listings. Given the fact that we're dealing with early listings, pricing for MSI's GeForce RTX 4080 Super 16GB, GeForce RTX 4070 Ti Super 16GB, GeForce RTX 4070 Super 12GB are rather high and total $1,461, $1,120, and $885 if converted to USD (without VAT).  

This retailer is not the only early bird to list Nvidia's upcoming RTX 40-series Super products — Austria-based MeisterSinger lists MSI's GeForce RTX 4080 Super Gaming X Slim 16 GB graphics card for €1,625 ($1,500 without VAT), which is significantly above the current prices of regular GeForce RTX 4080 boards.

Alleged Nvidia RTX 40-Series Super Specifications

*Specifications are unconfirmed.

Now that we are so close to the alleged formal launch date of Nvidia's new GeForce RTX Super graphics cards based on the Ada Lovelace architecture, we can be almost sure that the company is indeed about to introduce them, given listings by retailers and in customs databases. Yet, the specifications of these products should still be taken with a pinch of salt since they do not come from official sources.

Rumors about Nvidia's plans to refresh its desktop GPU lineup with GeForce RTX 40-Series GPUs with "Super" variants have been circulating for a while, but they did not seem credible. However, there are now several leaks alleging that Nvidia will introduce its GeForce RTX 4080 Super, GeForce RTX 4070 Ti Super, and GeForce RTX 4070 Super at an event on January 8, and then will gradually roll them out later next month.

The new GeForce RTX 40-series Super graphics cards are set to be officially introduced one day before CES 2024 takes off, according to IT Home, EXPreview, and Hassan Mujtaba, who published an alleged excerpt from an Nvidia document. Meanwhile, Nvidia's GeForce RTX 4070 Super is expected to hit store shelves on January 17; the GeForce RTX 4070 Ti Super is to be available on January 24; and the GeForce RTX 4080 Super is expected to hit the market on January 31.

There is another interesting part of these reports: re-emphasized specifications of Nvidia's purported GeForce RTX 4080 Super, GeForce RTX 4070 Ti Super, and GeForce RTX 4070 Super products, which will likely join the ranks of the best graphics cards, if they hit the market.

Alleged Nvidia RTX 40-Series Super Specifications

*Specifications are unconfirmed.

As it turns out, Nvidia's GeForce RTX 4080 Super is expected to come with a fully-enabled AD103 GPU and slightly faster memory than the predecessor, but it will maintain its total graphics power (TGP) of 320W. Interestingly, it is projected that the RTX 4080 Super will cost between $999 and $1,099, making the original RTX 4080 obsolete.

Meanwhile, the GeForce RTX 4070 Ti Super could come equipped with either AD103-275 or AD102-175 GPU featuring 8448 CUDA cores as well as 16GB of GDDR6X memory. It is rumored that the official recommended retail price is between $799 and $849, which will force graphics board makers to phase out non-Super GeForce RTX 4070 Ti fairly quickly — or significantly reduce its price.

As for the GeForce RTX 4070 Super, some believe that it will be based on the AD104-350 or AD103-175 GPU with 7168 CUDA cores, which is significantly more than 5888 CUDA cores in the case of the non-Super RTX 4070. Meanwhile, the product's MSRP is between $599 and $649, which will mean  the original GeForce RTX 4070 either becomes obsolete or gets a significant price cut.

While the information comes from multiple sources, it should be taken with the usual serving of salt since plans tend to change. Pricing, specs, and availability are not officially confirmed, though it's a relatively safe bet we'll see all three new GPUs next month.

Nvidia's GeForce RTX 4090D— the company's gaming graphics card designed specifically for Chinese market — is on track and its specifications are getting shape, according to recent leaks. Given the fact that the product has never been confirmed by Nvidia as well as the stance of the U.S. Department of Commerce regarding powerful GPUs for China, we can never be sure either about the specs or about the launch of the product at all. But TGP and clock leaks from @Zed__Wang and BenchLife.info we have some information. As ever, take leaks with a pinch of salt.

Nvidia's GeForce RTX 4090D is based on the AD102-250 GPU, and likely to have a base clock of 2280 MHz (up from 2230 MHz on the regular RTX 4090D) and posible boost clock of 2520 MHz (same as on the regular RTX 4090), according to @Zed__Wang, a renowned leaker. These clocks heavily suggest that the AD102-250 graphics processing unit will come with a reduced number of CUDA cores and other units, though we are speculating here. 

Now, with similar clocks and reduced number of CUDA cores, a GeForce RTX 4090D could match the performance of an RTX 4090 if overclocked (remember that the AD102 GPU was designed to be overclockable). So, to avoid this, the graphics card will come with total graphics power (TGP) capped at 425W (down from 450W in case of the original RTX 4090) and its overclocking capabilities will be locked, according to BenchLife.info.

The website reports that that Nvidia's add-in-board (AIB) partners are expected to receive samples of the AD102-250 GPU for testing this week, which means that we are going to see leaked specifications of the GeForce RTX 4090D shortly.

Now, while it is plausible that Nvidia is interested in making its $1599 graphics card available in China again,  and is willing to cut-down the performance of its AD102 below the threshold set by the U.S. Department of Commerce. It should be noted that the latter is unwilling to let Chinese have powerful GPUs that could be used for AI training or even technical computing at all. To that end, take any information about Nvidia's RTX 4090D or datacenter-oriented GPUs specifically tailored for China with a large grain of salt as the U.S. DoC could issue export rules and ban exports of these products to China even before their formal launch.

Nvidia's best graphics card for gaming, the RTX 4090 was recently struck with an export ban preventing the GPU from being sold to China. In response, Nvidia is reportedly developing a China-exclusive variant dubbed the RTX 4090 D (for "Dragon") — that's according to WCCFTech. The RTX 4090 D will supposedly feature lower performance and specifications to comply with U.S. export regulations. Sprinkle liberally with salt, obviously, though the idea absolutely makes sense.

The rumored GPU's specifications are unknown, but due to the US export guidelines, the RTX 4090 D will need to be less powerful than its outgoing counterpart. As a result, we will likely see the 4090 D coming with a lower TDP and a lower SM/core count to separate itself from the RTX 4090. Given how big the performance gap is between the RTX 4080 and the RTX 4090, this neutered 4090 could be quite a bit slower than the outgoing part (if the RTX 4080 Super doesn't come out). In a worst-case scenario, it might not even come with the AD102 GPU die.

The main metric that the 4090 D will need to meet is TPP, Total Processing Power. This is calculated by the maximum compute for a given bit-depth, using TFLOPS (or TOPS for integer work) multiplied by the number of bits. For the RTX 4090, TPP is 660.8 * 8 = 5,286 for FP8 work running on the Tensor cores (sparsity doesn't count). Also note that the value is the same for FP16: 330.4 * 16 = 5,286. The allowed limit is 4,800, so the RTX 4090 is about 10% "too powerful."

This is why the RTX 4080 can still be sold (and manufactured / assembled) in China. It has a TPP of 389.6 * 8 = 3,117. There's a secondary metric for data center parts called Performance Density (TPP divided by die area), but that doesn't come into play for consumer GPUs. It's the reason the Nvidia L4 also falls under the restrictions, as it has a PD rating of 6.6, using the AD104's 294 mm^2 die size and a TPP value of 1,936. (It's a good thing PD doesn't apply to consumer products, as the entirety of Nvidia's desktop RTX 40-series has a PD above 6.0.)

So what would Nvidia need to do for an RTX 4090 D to be allowable under U.S. export controls? Technically, anything below 4,800 TPP would suffice, but Nvidia will likely build in some wiggle room — to ensure overclocking as an example doesn't become a problem. Assume a clock speed of 2.7 GHz and we get a maximum number of SMs of 108. That would still have a TPP score of 4,778, however, so Nvidia will likely be aiming a bit lower, like around 100 SMs (4,428 TPP), just to be safe.

Making a neutered RTX 4090 is a smart move from Nvidia, as the Chinese market is huge, with a total population over 4x greater than the United States. As a result, it's worth Nvidia's time and effort to create a high-end replacement for the Chinese market that is similarly priced, so it can reclaim the ultra-high-end GPU market — 7900 XTX cards are reportedly doing well in China right now.

According to WCCFTech, the RTX 4090 D will be available from add-in board partners in China, similar to the RTX 4090, with no indication of a Founders Edition model. Pricing is also reported to be "similar" to the RTX 4090's old price in China — i.e. before the restrictions caused prices to shoot up — though it may be a tad less expensive to reflect its lower performance.

The demise of the RTX 4090 and the rise of the RTX 4090 D is the result of the U.S. Commerce Departments recent attempts to deny China access to advanced technologies. The RTX 4090 is an unfortunate casualty of the recent bans, though given what we're now seeing with RTX 4090 cards being retrofitted with blower-style  coolers, its understandable that even consumer parts may need restrictions.

The RTX 4090 D is one of several GPUs Nvidia is creating to comply with the U.S. regulations. Nvidia is also preparing several additional AI and HPC-focused GPUs that will comply with the United States' performance restrictions. The RTX 4090 D will reportedly launch in early 2024.

HandBrake Team this week released HandBrake 1.7.0 which adds support for hardware-accelerated AV1 transcoding on the latest graphics processors from AMD and Nvidia. As a result, HandBrake 1.7.0 now supports hardware-accelerated AV1 transcoding on all popular modern GPUs, including those from AMD, Nvidia, and Intel. In addition, the new version of HandBrake also quadruples the performance of SVT-AV1 encoding on Apple Silicon-based Macs thanks to CPU optimizations. 

Starting from version 1.7.0, HandBrake supports the AMD VCN AV1 encoder that is used by AMD's Radeon 7000-series graphics processors (RDNA 3), as well as the Nvidia 8th Gen NVENC encoder built into GeForce 40-series (Ada Lovelace) graphics chips. Both these GPU families are used in some of the best graphics cards. Previously HandBrake 1.6.0 gained support for Intel's Xe AV1 encoder supported by Xe and Arc-branded GPUs. Therefore, HandBrake can take advantage of hardware-accelerated AV1 encoding on all popular contemporary GPUs, except those integrated into Apple Silicon processors. 

HandBrake is one of the world's most popular transcoding programs. Its main advantage is that it supports virtually all widespread codecs on all hardware and software platforms. HandBrake's software-based SVT-AV1 is a very efficient encoder that supports all kinds of CPUs, which means that HandBrake can run on virtually everything. Meanwhile, hardware-accelerated processing is both faster and more power efficient, which brings loads of advantages, particularly on laptops running on battery power. 

Since version 1.7.0 is a major release for HandBrake, it brings loads of important improvements and a full list can be found over at GitHub. While improved support for HDR or added support for Apple VideoToolbox are significant enhancements in the newest version of the software, perhaps the most important tweak is the added support for drag and drop of multiple files at once on Apple macOS and Microsoft Windows-based systems. 

While the Nvidia RTX 4090 currently reigns as the fastest GPU among the best graphics cards, sitting in the pole position on our GPU benchmarks hierarchy, prices for it continue to climb. We wrote about the 10% increase since October just last week (the details are repeated below), but prices are now nearing the $2,000 mark on the cheapest cards. There's still a single Zotac RTX 4090 for $1,899 at Newegg, but the next cheapest is $1,999, and the next cheapest after that kicks the price up to $2,149. What exactly is going on?

As discussed previously, the RTX 4090 is now subject to China export restrictions starting November 17 — that's today. In short, any Nvidia card using the AD102 chip is now on the export list. Restricting these GPUs from export to China means that assembling 4090 graphics cards in China is also prohibited. Most graphics card companies do their card assembly in China for cost reasons, and so Asus, Gigabyte, MSI, PNY, etc. will all have to shift production of 4090 cards elsewhere — and China-centric brands may have to halt production completely.

There's also the fact that RTX 4090 inventory seems to be drying up. Many 4090 cards are out of stock, and we doubt that's due to increased demand. More probable is that Nvidia and its partners aren't making the cards anymore, or if they are, the costs to assemble the cards (see above) has increased significantly. Halting production on the 4090 would also make a lot of sense if something else is coming down the pipeline.

There's no official word from Nvidia, but the long-rumored RTX 4090 Ti (or perhaps RTX 4090 Super?) may actually come to market. We've been hearing supposed "leaks" about the 4090 Ti (sometimes also called the new Titan RTX) for most of the past year, including numerous shots of a massive quad-slot cooler. Nvidia left room for a higher performance AD102 part when it created the RTX 4090, but perhaps it scrapped the 4-slot design and will simply use the existing 4090 model with additional cores and cache enabled. Given the melting 16-pin connector snafu of the 4090, an RTX 4090 Ti with a 600W or higher TGP seems like asking for trouble, so Nvidia may also stick with a 450W TGP.

We can also cynically point out that Black Friday GPU sales are right around the corner. We don't usually see a lot of great deals on graphics cards, but we often see prices trend upward right before the sales kick off. That GPU that was selling for $600 earlier this year might suddenly jump to $700 for a month or so, only to "go on sale" for $600 again. "Save 14%!" But the above reasons mean that, out of all GPUs, the 4090 is least likely to see the usual price hike and then discount cycle this year.

We can't say for certain how each of the above factors into the RTX 4090 pricing equation, but we do know that 4090 prices have been creeping upward for the past few months. It's possible we'll get a new RTX 4090 Ti priced at the same $1,599 MSRP as the 4090 in the near future. That would be the best-case scenario, frankly. More likely, unfortunately, is that such a card — if it even exists and comes to market — will bump the MSRP up a notch. Those $1,999 RTX 4090 cards may soon become the base model RTX 4090 Ti / Super. Whatever the case, we'll continue to track prices as we wait to see where things end up.

A few weeks ago, we reported about a single-slot GeForce RTX 4060 Ti, one of the best graphics cards, with a blower fan allegedly coming from Colorful. To our surprise, a very similarly looking graphics card is listed by Chinese Galax, a prominent graphics card maker. This could mean the product could eventually end up in retail in Europe or the U.S.

As far as specifications are concerned, the Galax GeForce RTX 4060 Ti Unparalleled Max is just yet another RTX 4060 Ti product. It carries Nvidia's AD106-351 GPU featuring 4,352 CUDA cores and operates at the recommended Nvidia boost clocks of up to 2,535 MHz (5 MHz lower than on the original model). The GPU is mated with 16 GB of 18 GT/s GDDR6 memory over a 128-bit interface, which is, again, what Nvidia recommends. The card has four display outputs, comprising three DisplayPorts and one HDMI connector, making it quite versatile for various display setups.

The key selling point of the product is its form factor and dimensions. The board is 267 mm x 107 mm x 20 mm and fits easily into most computer cases, provided they have a spare eight-pin auxiliary PCIe power connector. Meanwhile, with a power rating of 165W, that blower fan may be pretty busy when the product is under high loads, so beware of noise.

It should be noted that Nvidia dislikes when its partners install blower-type fans on consumer products. To a large degree, this is because such graphics boards may pose competition to workstation-grade cards supplied by Nvidia and select partners.

One could fit multiple Galax GeForce RTX 4060 Ti Unparalleled Max boards into one system and use it for applications typically reserved for professional and server workloads, including video walls (which require multiple GPUs) and various AI/ML training and inference tasks that can use multiple FP16/BF16-capable GPUs running in parallel.

The saga continues over a year after the first problems with GeForce RTX 4090's melted 12VHPWR connectors were reported. A NorthridgeFix repairman claims he must fix about a hundred GeForce RTX 4090 graphics cards with failed connectors every month. He insists the connector failures are not user error, but issues with the connector design.

"We get about 20 to 25 [GeForce RTX] 4090 [graphics cards] a week," the repairman said in a NorthbridgeFix YouTube video."

This volume highlights just how common the issue is with this model. To illustrate how many GeForce RTX 4090 graphics boards come to repairs, the repairman said he had to get a half-mask respirator and an air purifier to protect his health when he fixes the melted connectors.

"We get them in so much that I bought myself this Hiroshima mask, so I do not have to smell the burn on those conductors," he said. "It smells like fireworks times ten. It cannot be healthy to keep smelling burnt connectors." 

Back in November 2022, Nvidia said that it was aware of 50 cases of melted GeForce RTX 4090 power connectors and that the standard issue was that they were not fully plugged into the graphics card, which essentially meant that it was a user error, not a failure of the connector.

Although Nvidia denies that the issues with melting power connectors on some of its GeForce RTX 40-series graphics cards lie in the peculiarities of the 12VHPWR auxiliary PCIe power connector itself, PCI SIG has now redesigned the power plug. The new 12V-2x6 addresses the main drawback of the 12VHPWR, its fixing force, and alters other aspects of the specification. This suggests that the 12VHPWR connector specification was so flawed that PCI SIG had to redesign it.

"So now it is safe to say the [GeForce RTX] 4090 melted connector is not a user problem," the repairman said. "We discussed this many times in the past, but I want to mention one last time that this is not a user error."

Edit, 11/12/2023: Corrected title.

According to renowned leaker @kopite7kimi, Nvidia's rumored GeForce RTX 40-series 'Super' graphics cards will see a performance boost, without an increased power consumption when compared to regular models. 

Since the information is from an unofficial rumor we must take it with a pinch of salt, as it may well prove to be inaccurate.

Nvidia's RTX 40-series Super family of graphics cards is expected to consist of three models: Nvidia RTX 4080 Super based on anAD103 GPU with 10240 CUDA cores, Nvidia RTX 4070 Ti Super featuring an AD103 GPU with 8448 CUDA cores, and Nvidia RTX 4070 Super powered by an AD103 or AD104 processor with 7168 CUDA cores. If kopite7kimi's information is correct, then these boards will consume 320W, 285W, and 220W of power, respectively. As a result, only the RTX 4070 Super will gain a 20W power draw.

Alleged Nvidia RTX 40-Series Super Specifications

*Specifications are unconfirmed.

Considering that Nvidia's RTX 40-series 'Super' products will likely offer noticeable performance gains compared to non-Super offerings, maintaining power consumption of the latter will be a major achievement of the GPU designer. We could easily see a Super card making it to our list of the best graphics cards.

Improving performance while keeping power consumption low is a key goal of chip design. Chipmakers and chip designers tend to engage in continuous process improvements (CPI) to increase production yields and lower performance variations by using statistical process control (SPC). Nvidia's AD103 and AD104 graphics processors have been in production for well over a year now. Both Nvidia and TSMC have likely discovered multiple ways to enhance Nvidia's custom 4N process technology to boost production yields, increase performance, and lower power consumption. Furthermore, Nvidia has also learnt better GPU power management, making it possible that it has also managed to increase performance of the new products and maintain power consumption of its old ones. 

Keep in mind that the information comes from an unofficial source, so take it with a grain of salt. After all, Nvidia has yet to confirm any GeForce RTX 40-series 'Super'-badged products coming.

The GeForce RTX 40-series, which has given us some of the best graphics cards, will reportedly receive the usual Nvidia "Super" treatment. @Kopite7kimi, a well-known leaker who tends to publish accurate information about Nvidia's upcoming graphics cards, has published possible specifications of Nvidia's GeForce RTX 4080 Super add-in-board as well as some other 'Super'-designated graphics cards. As it turns out, Nvidia is prepping a major GeForce RTX 40-series with at least three 'Super' cards.

Suppose the leaked information is correct, and it ought to have significant accuracy since the formal launch of Nvidia's GeForce RTX 40-series 'Super' lineup is 'weeks away,' according to the source. In that case, the GeForce RTX 4080 Super will indeed run the AD103 graphics processor in its complete configuration with all 10,240 CUDA cores enabled.

In addition, Nvidia allegedly plans the GeForce RTX 4070 Ti Super and GeForce RTX 4070 Super. The GeForce RTX 4070 Ti Super will supposedly get Nvidia's AD103 GPU with its 48 MB of L2 cache and a 16GB 256-bit GDDR6X memory subsystem, which will significantly increase its performance compared to the original RTX 4070 Ti model. Meanwhile, even the naming of the card looks odd.

Nvidia GeForce RTX 40-Series Super Specifications*

*Specifications are unconfirmed.

Even @kopite7kimi does not believe that GeForce RTX 4070 Ti Super and GeForce RTX 4070 Super will happen.

Meanwhile, considering colossal price and performance gaps in Nvidia's GeForce RTX 40-series lineup, an addition to the family is reasonable. Huge gaps were logical earlier this year to make people buy more expensive options, but at this point, it may make sense to fill these gaps and address gamers who are transitioning from the previous generation of graphics cards.

Yet again, the naming scheme of the supposed new graphics cards looks so off that we would take the information about them with a massive grain of salt.

When we first heard rumors about Nvidia's plan to add the GeForce RTX 4080 Super graphics card to its fleet, we considered it unlikely. The company feels pretty comfortable from a competition point of view and has not used the 'Super' brand for quite a while. But evidence mounts that the green company is prepping its GeForce RTX 4080 Super graphics board. 

The PCI ID for Nvidia's GeForce RTX 4080 Super graphics card was listed yesterday in a PCI ID repository. Realix, the developer of the popular diagnostic suite HWiNFO, declared the addition of GeForce RTX 4080 Super as an upcoming change for the next version of the program. Typically, diagnostic software developers receive information about forthcoming hardware well in advance, so we can consider this 'listing' as further proof that the RTX 4080 Super is incoming. This is not rock-solid proof, though, so take it with a grain of salt.

Nvidia's GeForce RTX 4080 Super is allegedly built on Nvidia's AD103 graphics processor, just like the regular version. However, the exact configuration of the GPU remains uncertain. It could operate with all of its 10,240 CUDA cores enabled, maintain the GeForce RTX 4080's 9,728 CUDA cores with increased clock speeds, or possess a different setup. In any case, it promises to be one of the best graphics cards around when it arrives. When it will be released is another question, though it would be logical for Nvidia to release it in time for the holiday season.

Speculating about the specifications of Nvidia's GeForce RTX 4080 Super is challenging at this stage. It is anticipated to be positioned above the standard GeForce RTX 4080 but below the top-tier GeForce RTX 4090. So, a minor performance uplift compared to its predecessor would suffice to sell it at a higher price since the gap between RTX 4080 and RTX 4090 is huge. Nvidia might also choose to release the GeForce RTX 4080 Super exclusively to OEMs, catering to PC manufacturers who prefer having new product versions annually to highlight improvements over previous models. 

In any case, Nvidia appears to be in the preparation phase for the product, which might end up a retail model or exclusively available as an add-in-board from selected PC manufacturers. Semi-official sources, while somewhat reliable, should not be considered conclusive as they are unofficial, and companies' plans, including product launches, are subject to change or cancellation.

Nvidia may be prepping mysterious GeForce RTX 4080 graphics card as a well-known PCI ID repository now lists such a product, which suggests that the device has been listed in a driver, Linux patch, GPU-Z validation, or BIOS. The product may end up as an OEM-only add-in-board available from select PC makers, but at least Nvidia seems to be prepping it, which somewhat corroborates the ongoing rumors that GeForce RTX 4080 Super is incoming. This is still a rumor and as such we must take the news with a pinch of salt.

Nvidia recently added 10de:2703 (NVIDIA_DEV.2703), an entry described as GeForce RTX 4080 Super to its R515 test driver, as noticed by StefanG3D from LaptopVideo2Go. The product is said to be based on Nvidia's AD103 graphics processor, though it is unclear whether the GPU comes with all of its 10,240 CUDA cores enabled, retains GeForce RTX 4080's 9,728 CUDA cores yet adopts higher clocks, or has another configuration. 

Now, while semi-official repositories are reliable sources of information, they are not the ultimate truth, not only because they are not official, but because plans can change and companies may cancel certain products because of their launches. The particular listing of the 10de:2703 device designated as GeForce RTX 4080 Super was added by T4CFantacy, who is a TechPowerUp GPU database with access to early GPU BIOS and GPU-Z validations, as noted by VideoCardz. Therefore, it is likely, but not definite that Nvidia and/or its partners are working on products called GeForce RTX 4080 Super. 

At this point it is hard to make guesses about specifications of Nvidia's GeForce RTX 4080 Super, though we would imagine that this one would sit above the regular GeForce RTX 4080 but below the GeForce RTX 4090, the best graphics card money can buy. If the unit offers a slight performance upgrade versus its predecessor, it will be good enough to further solidify its position on the market of desktop discrete graphics boards. Meanwhile, the company could make its GeForce RTX 4080 Super an OEM-only product aimed at PC makers who tend to want 'new' SKUs every year to show the difference between contemporary and last year's offerings. Anyhow, for now take the information about Nvidia's GeForce RTX 4080 with a grain of salt. 

In addition to the alleged GeForce RTX 4080 Super, the R515 test driver also lists H100 SXM5 96GB AI and HPC GPU for compute. This part will likely use high-density HBM3 or HBM3E memory and let applications access 96 GB of it, up from 80 GB accessible today.

Asus has quietly launched its RTX 4070 Ti TUF Gaming White OC Edition graphics card that brings together the enhanced reliability that TUF graphics cards are known for, advanced cooling, factory overclocking... and a white façade. The key selling point of the product is still its white outfit.

The Asus TUF Gaming GeForce RTX 4070 Ti White OC Edition graphics card is essentially the same as TUF Gaming GeForce RTX 4070 Ti OC Edition graphics card, but with a white cooling shroud and backplate. In fact, the white version of the product continues to use a black printed circuit board with a 9+4-phase voltage regulating module that is designed to ensure that Nvidia's AD104 GPU with 7680 CUDA cores gets enough of clean power to maximize its overclocking potential.

Whether dressed in white or black, the GPU can boost to 2760 MHz (which is slightly higher than 2610 MHz stock clocks set by Nvidia), while the 12GB of GDDR6X memory still runs at the standard 21 GT/s data transfer rate.

The add-in-board (AIB) gets power from a 12VHPWR connector that can deliver 450W–600W of power, but Asus doesn't disclose the actual power limit of the TUF Gaming RTX 4070 Ti. Meanwhile, the graphics card is equipped with a massive 3.25-slot three-fan cooling system that looks quite capable of cooling down a heavily overclocked GPU — even if this particular card only features a modest OC.

Keep in mind that end-user overclocking is a lottery (i.e. luck) and the GeForce RTX 4070 Ti is still one of the best graphics cards available, even at factory clocks.

One of the extra features of the TUF Gaming GeForce RTX 4070 Ti White OC Edition is its extra HDMI output. This AIB comes with three DisplayPort connectors and two HDMI 2.1 connector, which will be useful for those with two HDMI displays. Only four outputs can be used simultaneously, but the HDMI 2.1 ports have up to 42 Gbps of data throughput, compared to just 25.92 Gbps for DisplayPort 1.4a.

Like other modern graphics cards aimed at gamers seeking to create stylish builds, the Asus TUF Gaming GeForce RTX 4070 Ti White OC Edition is equipped with addressable RGB LEDs. Still, the lighting on TUF series cards tends to be more subdued than on Asus's ROG Strix line.

Since Asus introduced its first white GeForce RTX 40-series TUF board a few days ago, the AIB is not yet available and we cannot find its recommended price. Meanwhile, the regular (black) version of the product is priced at $849. The least expensive alternative using the same GPU right now is the Zotac RTX 4070 Ti Trinity OC that currently goes for $769.

The GeForce RTX 4080 is one of the best graphics cards, based more on performance rather than value. However, that may not be for long, as the GeForce RTX 4080 Super is reportedly on the horizon. New rumors from a distributor forum in China claim that Nvidia may be contemplating the discontinuation of the GeForce RTX 4080 if the GeForce RTX 4080 Super launches.

The unverified sources allege that Chinese retailers are stocking up on GeForce RTX 4080 graphics cards ahead of the rumored termination. Manufacturers have increased their shipments to push out as many units as possible. The results won't be positive if the GeForce RTX 4080 ends up on the guillotine, as its discontinuation could ultimately increase the pricing of the AD103-powered Ada Lovelace graphics card, though that also depends on how Nvidia prices the rumored 4080 Super.

The GeForce RTX 4080 Super's job would be to fill the gap between the GeForce RTX 4080 and the GeForce RTX 4090. The GeForce RTX 4080's 16GB of GDDR6X memory and 256-bit memory interface didn't cut it. Therefore, it's somewhat believable that the GeForce RTX 4080 Super could end up sporting 20GB of GDDR6X memory. It would be on the same level as the rivaling Radeon RX 7900 XT regarding memory capacity. Adding more memory to the GeForce RTX 4080 Super conveys an amplification of the memory interface, as the GeForce RTX 4080 Super will likely feature a 320-bit memory bus.

However, reputable hardware leaker kopite7kimi thinks differently. According to him, the GeForce RTX 4080 Super could be a simple refresh of the GeForce RTX 4080, comparable to what the GeForce RTX 2080 Super was to the original GeForce RTX 2080. In other words, the GeForce RTX 4080 Super might sport the full AD103 silicon with all 80 SMs and 10,240 CUDA cores — the regular GeForce RTX 4080 has 76 SMs. A complete AD103 die would grant the Super variant 5% more SMs, and boost clocks might be increased as well. The graphics card currently leverages 22.4 Gbps GDDR6X memory chips, but we could see 24 Gbps chips on the GeForce RTX 4080 Super, a 7% improvement in bandwidth.

The GeForce RTX 4080 has a $1,199 MSRP. While we've seen much speculation and rumor of the GeForce RTX 4080 Super's specifications, there's no word about pricing. Wishful thinking would hope to see Nvidia price the GeForce RTX 4080 Super at the same MSRP as the GeForce RTX 4080, or even lower, but that seems unlikely given the alleged specification bump.

Either way, many of us would agree that the GeForce RTX 4080 Super is what the GeForce RTX 4080 should have been. It would have helped Nvidia sell more GeForce RTX 4080 units. However, Nvidia was too busy with other ideas, like trying to push a gimped GeForce RTX 4080 12GB to the market.

Realistically, this second 4080 Super 16GB configuration would only be at best about 10% faster than the existing card. The 20GB alternative meanwhile might be as much as 25% faster by comparison. The AD103 model would also be the less expensive route, in theory, but that also depends on how many "unusable" AD102 and AD103 chips are currently floating around. Harvesting those chips with lower tier GPUs is what we typically see, whereas a fully enabled AD103 doesn't really offer much more than the existing part.

RTX 4070 Ti Super and RTX 4070 Super?

There's room for another Super GPU or two as well, which seem even harder to pin down. Only the RTX 4080 currently makes use of a nearly complete AD103 chip. We could get an RTX 4070 Ti upgrade slotting in between the 4080 and 4070 Ti, but what should Nvidia call it? 4070 Super doesn't make much sense — the RTX 2080 Super was slower than the existing 2080 Ti, ergo "Super" isn't quite as good as "Ti." Except then the GTX 1660 Super was basically equivalent in performance to the GTX 1660 Ti, so there are conflicting historical examples.

Calling it the RTX 4070 Ti Super would clearly spell out that this is an upgrade to the existing 4070 Ti — and it would be a longer and uglier name. Regardless of the name, Nvidia could potentially release a new card that fits below the existing 4080. Give it 66–72 SMs, but leave the 256-bit memory interface with 16GB of GDDR6X memory. Maybe clock the memory slower at 21 Gbps instead of the 4080's 22.4 Gbps if that helps to create a separation (depending on whether the new Super GPUs will replace or coexist with the existing cards). Price it at $799, and we're good to go, more or less.

There could even be one more card, an RTX 4070 Super, between the 4070 and 4070 Ti. There's not a huge gap, but both the 4070 and 4070 Ti use AD104, with 46 and 60 SMs, respectively. That's perhaps enough room to stuff in a 54 SM variant. But an RTX 4070 Ti Super and/or RTX 4070 Super both feel like a tough sell... which in retrospect isn't that far off what we felt about the 20-series Super cards.

That's the real heart of the matter. Nvidia has been making truckloads of money off its AI and data center chips lately, but the GeForce line isn't doing as well. There's a widespread sentiment that the 40-series is mostly a weaker generation, or at least a very poorly priced generation. If Nvidia has any interest in trying to regain some goodwill from gamers, giving us the option of higher performance cards at lower prices would be the best way to go about it.

As with all things in the rumors domain, we don't have any official word about any RTX 40-series Super parts. What we do know is that adding more model names to the lineup at higher prices isn't what people want. And this isn't 2021, where a mid-cycle price correction in the form of the RTX 3080 Ti and RTX 3070 Ti would be snapped up by hungry gamers. Quite the contrary: Most gamers seem more than ready to sit this generation out.

Will we actually get a 40-series refresh with Super cards? That remains to be seen, but if Nvidia does have some changes in the works, we can only hope they'll prove more attractive than the existing lineup.

Single-slot graphics cards are rare today. In fact, even low-end offerings like Intel's Arc A310 or AMD's Radeon RX 6400 tend to use dual-slot cooling systems. Single-slot graphics cards for gaming almost do not exist, but it looks like Colorful wants to change this and has developed a single-slot GeForce RTX 4060 Ti, which is one of the best graphics cards around. 

The GeForce RTX 4060 Ti 16 GB blower graphics card (discovered by @Zed_Wang) is powered by Nvidia's AD106 graphics processor with 4352 CUDA cores presumably running at Nvidia's recommended clocks of up to 2540 MHz. The board measures 267 mm x 107 mm x 20 mm and can be housed in almost any chassis that has an eight-pin auxiliary PCIe power connector. 

Another important aspect about this add-in-board (AIB) is that it is possible to install many of such graphics cards into one system, which is important for applications like video walls that use more than one monitors. Speaking of monitors, it is necessary to note that the single-slot GeForce RTX 4060 Ti has four display outputs: three DisplayPorts and one HDMI connector. In addition, multiple single-slot GeForce RTX 4060 Ti 16 GB could be used for FP16 or FP32 compute too.

Meanwhile, Nvidia does not really endorse consumer-grade graphics cards with blowers from its AIB partners as they compete against its own professional-grade cards that are sold at a huge premium. As a result, it is unclear whether the company will actually let Colorful, one of its major partners to release its GeForce RTX 4060 Ti 16 GB blower product commercially.

More rumors regarding the addition of some GeForce RTX 40 Super SKUs to Nvidia’s Ada Lovelace consumer graphics card portfolio have surfaced in recent days. The biggest potential news is that Chinese tech testing site Benchlife has seen signs of a GeForce RTX 4080 Super with 20GB VRAM on the way. Some may see this upgrade as a worthwhile improvement on the existing RTX 4080 with 16GB, especially if more CUDA cores are activated, too. Other key details about Super editions are even thinner on the ground, but we will try and distill the state of RTX 40 Super rumors below. Bring some salt.

Earlier this month, we mulled over the first worthwhile set of RTX 40 Super rumors which came via MEGAsizeGPU. The key morsels from that Tweet/X were that the purported GeForce RTX 4080 Super (maybe Ti) would arrive in early 2024. Moreover, it would be based on the AD102 GPU, have a TGP of under 450W, yet be priced similarly to the existing RTX 4080.

This week the same tipster followed up by asserting that Nvidia and partners were also preparing to release an RTX 4070 Super “based on AD103, has a 256bit bandwidth and 16G VRAM.” In addition, MEGAsizeGPU reckons there will be an RX 4070 D6 model with GDDR6. Benchlife actually embedded this tweet in its news section, before adding data from its own sources suggesting the GeForce RTX 4080 Super with 20GB is being prepared.

Table: Italics = rumor, Super model CUDA core counts unknown

With the rumored additions added to the Ada Lovelace series, we should end up with something like the above (ignoring lower-tier cards). GPU Rumorville has no idea about the CUDA core count for the new Super series GPUs. It was only asserted that the RTX 4060 D6 would be identical apart from the move from GDDR6X to GDDR6.

If the RTX 4080 Super does indeed come with the same price as the existing RTX 4080, that will probably precipitate a drop in the price of the non-Super card to help Nvidia products compete in this area of the market where AMD’s Radeon RX 7900 XTX flagship roams. Meanwhile the RTX 4070 Super and RTX 4070 D6 fill some more price / performance gaps that are wide enough to be targeted.

In light of the recent US / China sanction news, the RTX 4080 Super might also find it will have a special appeal in China. Perhaps the new AD102-based RTX 4080 Super card will have its specs tuned to come in just under the bandwidth / compute export red lines that come into force starting from Nov 16.

How To Choose a Graphics Card

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

CableMod introduced its original 12VHPWR angled adapter promising that it would solve overheating and melting issues surrounding Nvidia's adapter, but its aftermarket adapters appeared to also be prone to failures. Recently the company rolled out version 1.1 of its 12VHPWR angled adapters that adopt the new PCI SIG CEM 5.1 specification.

CableMod's 12VHPWR V1.1 (12V-2x6) angled adapter offers CEM5.1 compliance and is said to offer an improved friction-fit of the male connector which removes the wiggle between PCB and male connector. The device costs $39.90/€39.90 and is designed to solve issues with overheating and melting 12VHPWR dongles and adapters, both from makers of graphics cards and CableMod itself.

The updated PCI SIG 5.1 (12V-2×6) plug presents several modifications when compared to the original 12VHPWR connector. Firstly, the depth of the power terminal has been expanded from 4.2 mm to 4.45 mm. In addition, the sideband pin array's opening has been widened, moving from the previous measurements of 1.6 mm × 9.3 mm to 1.70 mm × 9.4 mm. One of the most notable changes, however, is the reduction of the sense pin length from 4 mm to 2.5 mm, ensuring that the power header is fully integrated with the connector. 

This adjustment is especially vital for safety reasons. If power connector is not adequately connected, the power supply avoids entering high power modes. This prevention mechanism stops potential overheating and possible melting situations, especially when a graphics card is drawing considerable energy. 

In response to the issues with the 12VHPWR connector, the PCI SIG set clear specifications for the new PCI SIG 5.1 (12V-2×6) design. They mandated a minimum current rating of 9.2A per pin, ensuring a temperature rise limit doesn't exceed 30 degrees Celsius above the ambient temperature at 12V DC. These connectors will bear an embossed H mark. Moreover, even if a single pin can manage a current of 9.2A, the entire assembly's current should not cross 55A RMS in either direction. The design also specifies the use of 16 AWG wires and pins and mandates that the connector can resist a pulling force of a minimum of 45.00N. 

One of the interesting things about the new 12VHPWR V1.1 (12V-2x6) adapter is that CableMod gives owners of the original adapter a $39.90/€39.90 coupon to buy themselves the new version of the device. Essentially, owners of the version 1.0 can get themselves a 12VHPWR V1.1 (12V-2x6) adapter for free.

Nvidia's partners have quietly started sales of Nvidia's RTX 5000 Ada Generation graphics card designed for professional visualization (ProViz) applications. The board carries the company's flagship AD102 GPU for client PCs, but in a severely cut down configuration to reduce its power consumption to 250W. Surprisingly, retailers are selling PNY's Nvidia RTX 5000 Ada boards at inflated prices.

The RTX 5000 Ada graphics card is now available from PC Connection, Ebay, ProVantage, ShopBLT, and ThinkMate, among other workstation-oriented retailers. Meanwhile, none of them currently sell this add-in-board at its MSRP of $4,000, with prices hovering 10–20 percent above this figure. ProVantage even goes so far as to suggest the list price for the card is $6,999 and that you're saving 37%. Nice try.

Nvidia's RTX 5000 Ada Generation graphics card is based on the AD102 graphics processor with 12,800 CUDA cores spread across 100 SMs. It offers peak compute performance of 65.3 FP32 TFLOPS and is mated to 32 GB of GDDR6 memory using a 256-bit interface. By contrast, Nvidia's range-topping RTX 6000 Ada powered by the same AD102 GPU boasts with 18,176 CUDA cores and delivers a whopping 91.6 FP32 TFLOPS. Furthermore, the RTX 6000 also has 48 GB of GDDR6 memory connected to the processor using a 384-bit bus.

Cutting down 30% of the GPU may seem a bit odd, but perhaps this was the only way to build a product that would sit below the RTX 6000 Ada and still deliver a noticeable, yet not dramatic performance gap with the top-of-the-line ProViz graphics card. Nvidia's full AD103 has a maximum of 10,240 CUDA cores spread over 80 SMs, so its potential performance would be far lower than a cut-down AD102.

Interestingly, despite severely cutting down the GPU configuration, Nvidia retained a rather high power rating of 250W for the part. As for the cooling system, the board is equipped with a dual-slot cooler with one blower. The board also uses a 12VHPWR PCIe auxiliary power connector.

Meanwhile, it remains to be seen how Nvidia might choose to leverage the AD103 chip for it's professional offerings. RTX 6000 Ada uses nearly the full AD102, with 142 (out of 144) SMs and the full 384-bit memory interface. RTX 5000 Ada uses AD102 with 100 SMs and only a 256-bit interface. Then there's a big jump down to RTX 4500 Ada with a full AD104 implementation of 60 SMs and a 192-bit interface. RTX 4000 Ada and RTX 4000 SFF Ada wrap things up with a partial AD104 chip using 48 SMs and a 160-bit interface.

There's room for an RTX Ada card between the 4500 Ada and 5000 Ada, in other words, using AD103. However, Nvidia hasn't announced such an offering yet, and there's no room for it within the standard naming structure (i.e. RTX 5500 Ada would be this card). Perhaps Nvidia will simply bypass using AD103 on professional cards entirely.

MSI has quietly introduced its new Ventus Essential series of graphics cards that will complement the company's regular Ventus family of boards (as noticed by @momomo_us). The new Ventus Essential cards will offer almost exactly the same specifications as normal Ventus add-in-boards, but will likely have a different bill-of-materials (BOM), which will make MSI's costs slightly lower.

MSI's Ventus Essential lineup currently consists of the GeForce RTX 4070, GeForce RTX 4070 Ti, and GeForce RTX 4090, which rank among the best graphics cards. All boards have the same clocks and cooling systems as standard Ventus cards, but Ventus Essential features two DisplayPort and two HDMI outputs, whereas regular Ventus boards have three DisplayPort and one HDMI output. Essential obviously points to something basic, but this may not be the case.

"It has slight structural improvements over Ventus but mostly the same design/specs," wrote Hassan Mujtaba of WccfTech. Note that MSI does not officially disclose any differences between its Ventus and Ventus Essential graphics boards, so let us try to speculate a little.

The display outputs are an interesting change. Right now, DisplayPort on RTX 40-series cards remains on the DP1.4a spec, which has a maximum data rate of 25.92 Gbps using 8b/10b encoding. HDMI 2.1 meanwhile offers up to 42 Gbps data rates using 16b/18b encoding. That's 62% more bandwidth, which means HDMI 2.1 can output up to 8K 120 Hz using DSC (Display Stream Compression), while DP1.4a tops out at around 8K 72 Hz using DSC. But HDMI typically costs more for licensing than DisplayPort, which makes this particular change a bit odd.

Besides the video ports, modern graphics cards are highly complex devices that use hundreds of components. Custom graphics cards, such as MSI's Ventus series, tend to slightly change the bill-of-materials over their lifetimes because some components may get cheaper and others may disappear from the market. But it looks like MSI wants to concurrently offer Ventus and Ventus Essential parts, possibly with slightly different bundles and slightly different BOMs.

Such an approach allows MSI to be flexible with its costs and offer its GeForce RTX 40-series products at more competitive prices. Of course, it remains to be seen whether these Ventus Essential boards will be significantly cheaper than non-Essential Ventus cardsbut the key point is that with more SKUs the company will be able to pick the right series and produce it if it finds it more suitable.

Samsung's gigantic 57-inch Odyssey Neo G9 display is certainly one of the most anticipated gaming monitors of the year. Impressively, it supports an ultrawide resolution of 7680x2160 at a refresh rate of 240 Hz. But it appears that the gamer's dream display cannot be handled by one of the world's highest-performing gaming graphics cards, namely the Nvidia GeForce RTX 4090. Whether using DisplayPort 2.1 or HDMI 2.1 connection, some limitation seems to be present, a review of the display by QuasarZone revealed.

An Odd Limitation

For now, the only graphics cards that can handle a 7680x2160 resolution at a refresh rate of 240 Hz with Samsung's curved 57-inch Odyssey Neo G9 are AMD's Radeon RX 7000-series that support a DisplayPort 2.1 with UHBR 13.5 (and offering a 54 Gbit/s of bandwidth). Meanwhile, the review by QuasarZone indicates that AMD's Radeon RX 7000-series boards can also support 7680x2160 at 240 Hz using HDMI 2.1 when used with a proper 48G cable. By contrast, neither Nvidia's GeForce RTX 40-series GPUs nor Intel's Arc A700-series GPUs could support this resolution at this refresh rate using any kind of connection, according to QuasarZone.

Such a limitation is very odd, and its reasons are unknown. However, there could be restrictions in Intel and Nvidia GPU display pipelines or certain limitations in firmware or software. 

Possible Reasons

Samsung's curved 57-inch Odyssey Neo G9 display is essentially two 4K 240 Hz displays glued together. Transferring an uncompressed 4K stream with 10-bit per color at 240 Hz requires 54.84 Gbit/s of bandwidth, and transferring two such streams requires 109.68 Gbit/s of bandwidth. The highest bandwidth one can get today from a single display connection is 77.37 Gbit/s (using DisplayPort 2.1 with UHBR 20) or 48 Gbit/s (using HDMI 2.1), so whatever connection is used, Display Stream Compression (DSC) technology is required. 

DSC can compress any image to 8 bits per pixel (bpp) without a perceptual loss in image quality. This results in a 3:1 compression ratio for a 24 bpp image or a 3.75X compression ratio for a 30 bpp image, according to Rambus. The exact compression ratio achieved can vary based on the content: for visually complex segments, DSC might use a lower compression ratio, whereas, for simpler segments, it can compress more aggressively. Meanwhile, 3:1 is a reasonable compression ratio to expect. With DSC enabled, Samsung's Odyssey Neo G9 theoretically only needs 36.56 Gbit/s, and 48 Gbit/s provided by HDMI 2.1 should suffice. 

But it may not be this easy. Before the compression process starts, the uncompressed video frame is first divided into fixed-size segments or slices. One of the configuration parameters in DSC is the slice width. Combined with the total resolution, this parameter will determine how many slices a frame is divided into. While DSC itself is flexible, the implementation of DSC in certain hardware or software might be optimized for popular resolutions, like 1080p, 1440p, and 4K, and may not be optimized for dual 4K. If a particular resolution results in an inconvenient or non-optimal number of slices, it might be less efficient or potentially problematic for certain implementations. The hardware often processes multiple slices in parallel, requiring multiple DSC pipelines. If the number of pipelines is not enough for a particular DSC implementation for a particular resolution, a problem might occur. 

For now, we can only speculate why modern GPUs cannot handle Samsung's curved 57-inch Odyssey Neo G9 display properly and whether it can be fixed with software and firmware updates. If QuasarZone's findings are correct, this monitor can only work properly with AMD's latest Radeon RX 7000-series graphics cards.

The world's leading chip design firms experienced a significant revenue boost because of the ongoing generative AI frenzy. They reached a collective $38.1 billion, a 12.5% rise from the previous quarter, according to TrendForce. Nvidia, which sold $10 billion worth of datacenter hardware emerged as the highest-earning fabless chip designer, overtaking Qualcomm in the rankings.

Nvidia capitalized on the global surge of artificial intelligence services based on large language models and saw its data center earnings soar by 105%. This growth was further fueled by the successful launch and shipment of their cutting-edge Hopper and Ampere architecture HGX systems. As a result, Nvidia's Q2 earnings reached a staggering $11.33 billion, marking a 68.3% increase quarter-over-quarter, TrendForce reports.

On the other hand, Qualcomm faced challenges in Q2. The Android smartphone market declined, which led to a dip in the company's revenue (although Apple's pre-purchases of modems helped a bit). Qualcomm reported Q2 earnings of approximately $7.17 billion, reflecting a 9.7% decline quarter-over-quarter.

Broadcom, another major player, ended the quarter with a revenue close to $6.9 billion, which was in line with its performance in the previous quarter. While the company did benefit from the AI-driven demand, particularly for high-end switches and routers, they also encountered challenges in sectors like server storage and broadband.

AMD, which is now a major supplier of server CPUs, reported a Q2 revenue of around $5.36 billion. The company's performance was impacted by a decrease in gaming GPU and embedded chips sales. Meanwhile, MediaTek experienced a revival, with its Q2 earnings climbing to a solid $3.2 billion, thanks to components like TV SoCs and Wi-Fi. 

As the industry looks ahead to the third quarter, there's cautious optimism, according to TrendForce. While inventory levels seem promising, there's a noticeable dip in end-user demand. However, the increasing interest in AI and large language models from major companies suggests potential for continued growth, with projections hinting at another double-digit rise for the top IC design firms.

Asus has finally officially introduced its flagship graphics card: the ROG Matrix GeForce RTX 4090. The board comes with the highest out-of-box AD102 GPU clocks, ultimate tweaking ability, a sophisticated voltage regulating module (VRM), an advanced cooling system, and extreme overclockability. Asus has plenty of reasons to suggest this is the world's best graphics card — and it's got a price to match. The ROG Matrix RTX 4090 carries a recommended price tag of $3,199, which is twice Nvidia's GeForce RTX 4090 MSRP of $1,599.

The Asus ROG Matrix GeForce RTX 4090 is based on the AD102 GPU with 16,384 CUDA cores that come with a 2700 MHz maximum frequency out-of-box (up from 2520 MHz, recommended by Nvidia). These CUDA cores are mated with 24GB of 21 GT/s GDDR6X memory. The graphics card relies on a custom printed circuit board (PCB) with a 24-phase VRM with high-current power stages and one 12VHPWR connector that can deliver up to 600W of very clean power to the GPU. The PCB has numerous sensors, which allow it to monitor the temperature of the GPU, memory chips, VRM, coils, and even PCB itself. 

The graphics card features an all-in-one cooling system with a custom-engineered cold plate that cools the GPU, memory, and VRM, as well as an embedded pump, 700mm of tubing, and a 360mm radiator. To maximize heat transfer from the graphics processor to the cooler, Asus uses liquid metal thermal interface material (TIM) — an industry first for a GPU. The cooling system also has magnetic, daisy-chainable fans with RGB lighting, which allow for minimal cable clutter.

Premium hardware isn't the only strength of the Asus ROG Matrix RTX 4090. The manufacturer also amplified its GPU Tweak III software with more monitoring and overclocking features that take advantage of the card's extra capabilities and sensors. The program allows the user to set a power target, increase GPU voltage, set the GPU boost clock, adjust the voltage/frequency curve, modify the memory clock, and change the fan speed. Meanwhile, the Thermal Map capability provides real-time temperature readouts across the card, while the new mileage feature tracks the card's usage at different power levels.

An interesting feature is that the card supports Power Detector+, which reads the 12VHPWR connector and measures currents across all power wires to detect any anomalies (advising the user to reseat the cable, if necessary).

"The ROG Matrix GeForce RTX 4090 is an overclocker's dream," says a statement by Asus. "In fact, since this card was unveiled at Computex, it has claimed three World Records and five Global First Place records for a total of seven overclocking records in 3DMark11 Performance, 3DMark Fire Strike Extreme, Unigine Superposition 1080p Xtreme, 3DMark Time Spy Extreme, 3DMark Port Royal, GPUPI v3.3 1B, GPUPI v3.3 32B, and Unigine Superposition 8K Optimized."

Indeed, without any modifications, Asus' ROG Matrix GeForce RTX 4090 broke the 4 GHz GPU clock barrier earlier this year, so the company isn't wrong — this board does appear to be an overclocker's dream. But while it's certainly faster than any other RTX 4090 currently on the market, the ROG Matrix GeForce RTX 4090 also costs twice as much as a regular RTX 4090. It'll probably be hard for many to justify spending an extra $1,600 on the extra performance, even if they're into overclocking.

The Asus ROG Matrix RTX 4090 is, obviously, not for everyone. It's a limited-edition product aimed at die-hard overclockers who are looking to squeeze every last drop of performance out of their hardware. It's for people who actually enjoy tweaking their systems to gain a couple of points on a benchmarking test. Naturally, that also makes it perfect for bragging rights.

Asus is set to launch its flagship ROG Matrix GeForce RTX 4090 graphics card next week on September 19, 2023. The graphics board was announced at Computex, but the company only shared that it uses an innovative closed-loop cooling system with a liquid metal thermal interface. Meanwhile, the product has already become a legend as it nearly conquered a 4 GHz GPU frequency, an industry's first.

"The wait is almost over," an Asus tweet reads. "Count down to the official launch of our flagship graphics card – the ROG Matrix GeForce RTX 4090."

The Asus ROG Matrix GeForce RTX 4090 is one of the most mysterious products introduced recently. We know it carries Nvidia's AD102 graphics processor and 24 GB of GDDR6X memory. However, Asus has been tight-lipped about out-of-box clocks of both the GPU and SGRAM, instead focusing on its custom printed circuit board with a sophisticated voltage regulating module and its cooling system with Liquid Metal thermal interface.

Plenty of clean power and a powerful cooling system are indeed required for successful overclocking. So, all makers of graphics cards tend to use custom PCBs with powerful VRMs and an all-in-one 360-mm liquid cooling system. What sets the ROG Matrix RTX 4090 graphics card apart is its use of a special ingredient: liquid metal thermal interface material (TIM), which enhances heat transfer from the hot components to the cooling system. While Asus doesn't disclose the specific type of liquid metal TIM employed, such materials are typically composed of gallium or gallium alloys known for their exceptional heat conductivity.

However, utilizing liquid metal thermal interfaces presents certain risks and challenges. These TIMs are electrically conductive, potentially leading to short circuits if not handled with care and precision. Moreover, they can be corrosive to specific metals like aluminum. Despite these challenges, Asus asserts its experience with using liquid metal TIMs in laptops for several years, indicating confidence in their application for graphics cards.

In any case, we already know that the Asus ROG Matrix GeForce RTX 4090 can hit nearly 4 GHz GPU with liquid nitrogen, so we'll learn next week what it is capable of when equipped with its stock cooling system and Liquid Metal TIM. Asus clearly wanted to build the best graphics card possible, so we will find out soon enough whether it has succeeded.

AMD's Radeon RX 7800 XT may not be radically better than its predecessor in terms of performance, but it is enough to challenge Nvidia's GeForce RTX 4070, which is perhaps why some graphics card manufacturers and retailers are cutting prices of custom GeForce RTX 4070 models by $50 in the U.S.

Currently, Amazon.com, BestBuy, B&H Photo Video, and Newegg have GeForce RTX 4070 graphics cards available for $549 from Asus, Gigabyte, and Zotac. The list includes several models:

• Asus GeForce RTX 4070 Dual at B&H Photo Video.
• Gigabyte GeForce RTX 4070 WindForce OC at Amazon, Best Buy, B&H Photo Video, and Newegg.
• Zotac Gaming GeForce RTX 4070 Twin Edge OC Spider-Man: Across The Spider-Verse at Amazon, and Newegg.

Nvidia's GeForce RTX 4070 carries an MSRP of $599, but AMD's Radeon RX 7800 XT is priced at $499. Despite being priced lower, AMD's offering sometimes outperforms its more expensive rival in many games that depend on rasterization, with performance that's close to the competitor in some ray tracing-supporting games as well. This certainly increases the value of AMD's product, earning it a spot on our list of the best graphics cards.

We suspect that's why at least three GPU manufacturers are partnering with select retailers to offer certain GeForce RTX 4070 models with a $50 discount. However, it's unknown whether this is a short-term campaign or if other manufacturers and retailers will join it and keep the base price at $549 for the long-term.

Regardless, it doesn't look like Nvidia is quietly reducing the recommended price of its GeForce RTX 4070 graphics cards. The official MSRP remains at $599, just as the RTX 4060 Ti sits at $399 with the 4060 Ti 16GB model at $499. Now that AMD's Navi 32 GPUs have arrived, you can now find all three of those RTX cards for less than their respective MSRPs.

After it turned out that Nvidia-supplied 16-pin 12VHPWR adapters could melt if inappropriately connected to the graphics card, many GeForce RTX 40-series users opted to use aftermarket 12VHPWR adapters, which are also prone to failure if improperly fitted. Now, it appears that MSI is refusing to replace a graphics card that failed because of the CableMod dongle.

A Reddit user has published his purported communications with MSI, and the company's employee claims that their graphics card failed because of a 12VHPWR CableMod adapter with 'manufacturing and tolerance issues,' which voids the MSI warranty.

"There is an ongoing issue with CableMod's adapter seemingly due to manufacturing and tolerance issues with the CableMod adapter," an alleged statement by MSI reads. "You will need to contact CableMod as their third-party adapter has caused an issue and voided the warranty on your graphics card."

CableMod has neither confirmed nor denied the possibility that a graphics card can fail because of its 12VHPWR adapter, but has promised to look into the issue.

"Sorry to hear that, please put in a ticket with our support team," Matt from CableMod said. "They will make sure you get covered for sure in full."

After examining approximately 50 cases of 16-pin 12VHPWR cable adapters overheating and melting, Nvidia identified a recurrent problem: the connectors were not securely attached to the graphics card. The firm advises connecting the dongles before mounting the graphics card to the motherboard. However, repeated issues with cards deployed in the market imply that the cables could potentially work their way loose over time. 

Ensuring power connectors are securely fitted is crucial, as almost all cables are susceptible to overheating when loosely connected. Additionally, it is imperative to securely attach the connector to Nvidia's range-topping GeForce RTX GPUs. The top-tier 40-series GPUs can pull as much as 450W from the PSU, demanding top-notch cables and stable connections to manage such a significant power transfer.

Nvidia's codenamed Blackwell family of graphics processors will contain five different GPUs and will lack a direct successor to Nvidia's highly-successful AD104 chip, according to leaks by Chiphell and kopite7kimi (via VideoCardz). The information is unofficial and may be inaccurate, but if true, Nvidia will have to address market segments currently addressed by AD104 with two different GPUs.

Apparently, Nvidia's Blackwell family of graphics processors contains five chips codenamed GB202, GB203, GB205, GB206, and GB207. Nvidia's Ada Lovelace family released to date also contains five processors (AD102, AD103, AD104, AD106, AD107), just like the company's Ampere lineup of GPUs (GA102, GA103, GA104, GA106, GA107) that still powers some of the best graphics cards. Meanwhile, Nvidia's Turing family comprised of three members, whereas the Pascal lineup contained five GPUs. 

It is unclear why Nvidia's Blackwell family is said to feature GB200-series GPUs, but not GB100-series graphics processors. Typically, 200-series represents re-spinned GPUs.

Historically, Nvidia's XXX04 served performance-mainstream segment of the market and contained 50% - 66% transistors of the top-of-the-range part. The gap between the high-end and the performance mainstream part was quite noticeable. To fill the gap between its GA102 and GA104 GPUs in the Ampere era, Nvidia introduced GA103 part and did the same with the Ada Lovelace family: there is the AD103 sitting between the AD102 and AD104.  

While the now Nvidia's lineup no longer has wide gaps, a cut-down AD103 would overlap with the full AD104, which means that Nvidia has to either throw away AD103 GPUs that have one or more defective streaming multiprocessors or even CUDA cores so as not to compete against AD104, or keep them and then use them quietly to substitute AD104, which means cutting them down substantially and not using the whole potential of AD103.

Apparently, the company wants to avoid such a situation in the future. As a result, its GB202 will keep addressing the highest-end of the market (e.g., GeForce RTX 5090, GeForce RTX 5090 Ti), its GB203 will address high-end and performance-mainstream segments (e.g., GeForce RTX 5080, GeForce RTX 5070 Ti, and GeForce RTX 5070), while GB105 will address mainstream part of the market (GeForce RTX 5060 Ti, GeForce RTX 5060). This will enable Nvidia to use all GB203 silicon that it has even with some defective SMs. Of course, we are speculating here.

Nvidia's next-generation Blackwell GPUs are expected to hit the market in late 2024 or early 2025, so the company's plans may change a lot between now and then. Therefore, take this information with a grain of salt for now.

Asus has started to sell some of the best graphics cards under the company's Tianxuan (Heavenly Chosen) portfolio. Marketed under the TX Gaming moniker, the new trio of graphics cards encompasses the GeForce RTX 4070, GeForce RTX 4060 Ti, and GeForce RTX 4060.

The Tianxuan series may be new to some since it's one of Asus' China-exclusive sub-brands. The company has launched many gaming laptops from the Tianxuan series but is extending the product portfolio to graphics cards, specifically the GeForce RTX 40-series parts.

The TX Gaming graphics cards spit images of Asus' Atlas Shark (Megalodon) series. They're using the same cooler but with a different color theme. While the Atlas Shark series cooler stick to a black and grey combination, the TX Gaming series arrives with a white and grey theme with teal accents. There aren't any visible changes to the cooling solution. It's still a thick shroud with three axial, dual-ball bearing cooling fans. The graphics cards also have a reinforced metal backplate that helps with cooling.

Regardless of the model, the TX Gaming graphics card measures 12.48 x 5.35 x 2.13 inches (317 x 136 x 54mm), so we're looking at a triple-slot design. The size of the cooler seems a bit overkill for the lower-tier SKUs, such as the GeForce RTX 4060 Ti (160W) or GeForce RTX 4060 (115W). However, it looks like Asus didn't want to waste resources redesigning the cooler for each SKU. Instead, the company used the same cooler for all three graphics cards.

As their part numbers insinuate, the TX Gaming graphics cards feature a factory overclock. However, Asus didn't share this information on its product pages or online store. We only know that the graphics cards require one 8-pin PCIe power connector for external power. Many will be glad that Asus didn't opt for the pesky 16-pin power connector. At any rate, you'd want a power supply with a minimum capacity of 650W for whichever model you pick up. Asus didn't alter the design of the display outputs. Therefore, the TX Gaming graphics cards provide one HDMI 2.1 port and up to three DisplayPort 1.4a outputs.

Asus sells the TX Gaming GeForce RTX 4070 (TX-RTX4070-O12G-Gaming) for $733.54. Meanwhile, the TX Gaming GeForce RTX 4060 Ti (TX-RTX4060TI-O8G-Gaming) and TX Gaming GeForce RTX 4060 (TX-RTX4060-O8G-Gaming) retail for $526.91 and $386.82, respectively.

Nvidia has introduced three high-performance professional graphics cards based on the Ada Lovelace architecture for workstations as well as a server-grade grade professional board that can be used both for remote graphics and light AI applications. The introduction completes transition of Nvidia's ProViz family of high-performance products to its latest Ada Lovelace architecture.

To address performance-demanding professional graphics applications, such as computer aided design and digital content creation, Nvidia add three new products: the RTX 4000 20GB, the RTX 4500 24GB, and the RTX 5000 32GB boards based on the Ada Lovelace architecture. In addition, Nvidia is rolling out its L40S datacenter board with 48GB of memory.

The new Nvidia RTX 4000 20GB workstation graphics card largely mimics the RTX 4000 SFF product released several month ago, but it uses a full-height PCB albeit comes with a single-slot cooling system and is rated for 130W. The part is powered by the AD104 GPU with 6144 CUDA cores that is clocked higher compared to the SFF variant and thus delivers up to 26.7 FP32 TFLOPS of compute throughput, which is comparable to compute performance of Nvidia's GeForce RTX 4070. This board will offer higher performance than the RTX 4000 SFF for the same price of $1,250 in September.

The green company is also rolling out its Nvidia RTX 4500 24GB featuring the AD104 GPU with 7,680 CUDA cores that offers up to 39.6 FP32 TFLOPS of compute performance, which is on par with the GeForce RTX 4070 Ti. The ProViz graphics card is equipped with a dual-slot cooling system with a blower fan and is rated for up to 210W of power. The product is set to be available in October for the price of $2,250.

Yet another graphics cards that is being rolled out today is the Nvidia RTX 5000 32GB based on the severely cut-down AD102 graphics processor with 12,800 CUDA cores that delivers compute performance of 65.3 FP32 TFLOPS. This unit is positioned to sit below the flagship RTX 6000 Ada and the whopping performance difference between the two parts implies that over time Nvidia might offer a solution that will sit between these models. In the meantime, Nvidia will have its RTX 5000 32GB for $4,000 and RTX 6000 48GB Ada for $6,800.

The new workstation boards will be used by companies like Boxx, Dell, HP, Lenovo, and Lambda in their upcoming workstations this fall. In addition, these boards will be available from Nvidia's resellers, such as Arrow and Ingram from such AIB suppliers as Leadtek, PNY, and Ryoyo,

But as there are professionals who use remote workstations, Nvidia is also rolling out its L40S Ada datacenter card that uses the AD102 GPU with 18,176 CUDA cores that delivers a whopping 91.6 FP32 TFLOPS, which is in line with performance of the RTX 6000 Ada. The L40S Ada will be first used in Nvidia's OVX servers used for graphics AI, and video processing, but eventually they will likely end up in different machines as well. While the L40S Ada is clearly a datacenter product with a passive cooling solution, it still has display outputs, so it can be installed into a workstation assuming that there is enough airflow inside or a special blower attached to the board. 

"As generative AI transforms every industry, enterprises are increasingly seeking large-scale compute resources in the data center," said Bob Pette, vice president of professional visualization at NVIDIA. "OVX systems with NVIDIA L40S GPUs accelerate AI, graphics and video processing workloads, and meet the demanding performance requirements of an ever-increasing set of complex and diverse applications."

Gigabyte has quietly launched a new GeForce RTX 4060 graphics card with a low profile (LP) design. We first caught a glimpse of this sleek and compact RTX 4060 in early July. Now we have the full details and specifications and can be quite confident this product will be on shelves soon. HTPC and compact PC upgraders and DIYers may be particularly interested in this card.

From Nvidia’s existing range of Ada Lovelace GPU-based products, the RTX 4060 is the best suited to cramped form factors due to its thrifty 115 W TDP, the lowest power consumption in the family until some RTX 4050 cards come along. Mini-ITX RTX 4060 designs from Nvidia AIBs are quite common already, but LP cards are much thinner on the ground. This is actually Gigabyte’s first Ada Lovelace LP product.

We have seen Gigabyte’s previous generations feature LP designs, but this Gigabyte GeForce RTX 4060 OC Low Profile 8G (GV-N4060OC-8GL) is the first outing for what seems like a new triple fan cooler. Gigabyte says this new custom design is 182mm long, 40mm thick, and 69mm tall (LP bracket) and features “pure copper heat pipes that make direct contact with the GPU.” The heat pipes are threaded through a heat sink, and cooled by the trio of fans.

In nearly all the gallery and promotional images of this graphics card, the full-height bracket is affixed to the card, but Gigabyte says there is an LP bracket in the box. We didn’t really doubt it, but it is good to see it confirmed. As per the previous leak, the connectivity provided allows for up to four simultaneous displays via two DP and two HDMI ports.

From the name of this graphics card model, you will be expecting a factory overclock, but you aren't going to be thrilled by the speed. According to Gigabyte, this LP card’s GPU runs up to 2,475 MHz out of the box, which is a mere 15 MHz better than the reference card. All other technical specifications of this 3,072 CUDA core and 8 GB of GDDR6 memory graphics card follow the reference. Check out our GeForce RTX 4060 review for the performance levels this LP card should be capable of delivering.

So, the Gigabyte GeForce RTX 4060 OC Low Profile 8G is very much a ‘known quantity’ as it launches. However, potential buyers should wait for third-party reviews to hear about the triple fan cooler’s noise profile (smaller fans can be more irritating). Also, there might be a price premium for this compact design.

Nvidia's prototype quad-slot cooling system for never-released 40-series graphics card has made quite a splash among enthusiasts in the recent months partly because of an unusual PCB mounting and partly because of its size. But the cooler apparently has an interesting wrinkle: it has a third fan inside to boost its performance even further.

 @Harukaze5719 has found more pictures of the ultimate GeForce RTX 40-series cooling system published by Hayaka, the owner of the cooler who once wanted to sell it for $120,000. This time around, Hayaka disassembled the cooling system and discovered that in addition to two large fans, it has a small fan inside that is meant to generate additional airflow. It also turned out that the prototype cooling system has 22 heat pipes to effectively distribute heat dissipated by the AD102 graphics processor as well as its voltage regulating module.

PCB placement, 22 heat pipes, and three fans indicate that the prototype of the monstrous graphics card was meant to deliver maximum possible performance at maximum possible power. To that end, we are dealing with a prototype cooling system that is meant to deliver overkill performance for simply the best graphics card available.

Meanwhile, an interesting thing about the current prototype is that it only has one 12VHPWR auxiliary PCIe power connector that can deliver up to 600W of power, so a cooler this large might be an overkill for the board. Then again, if Nvidia's plan was to develop the world's highest-performing air cooler for an ultimate graphics card, then this monstrous unit was a way to go.

Yet, it would be interesting to literally hear a cooling system with three fans one of which is located inside the cooler. While Nvidia's cooling systems for Founders Edition graphics cards tend to be rather quiet, it is unlikely that this particular device aimed at the graphics cards sometimes referred to as 'The Beast' was ever meant to be quiet.