Search the Community

Remains pricier than past xx70-tier offerings, but the performance bump is nice. The GeForce RTX 4070 Super Founders Edition. Andrew Cunningham Judging by the comments on YouTube reviews, you'd think Nvidia's RTX 4070, launched in April 2023 for $599, was a terrible graphics card. The reaction wasn't as brutal as it was among commenters and reviewers for the 4060 Ti a month later (a "waste of sand," declared Gamers Nexus), but you'd usually find praise for its power efficiency but criticism of its price (high for a xx70 card) and its performance improvement relative to the previous generation (only about as fast as an RTX 3080, sometimes less). Those are all largely valid criticisms. But the 4070 is Nvidia's most popular RTX 4000-series desktop GPU, at least according to the (admittedly flaky and opaque) Steam Hardware Survey data for December 2023. It's not in the top 10—this is dominated by older midrange GeForce cards that have been out a lot longer—but it's doing better than Nvidia's other 40-series desktop cards and better than every one of AMD's RX 7000-series cards put together. The release of the RTX 4070 Super should help bring the enthusiast commentariat version of reality and the Steam Hardware Survey's version of reality into closer sync with one another. At the same $599 price—still more than the $499 of the 2070 Super or 3070, but not higher than the 4070—you get performance that Nvidia says is more in line with the RTX 3090. And the power efficiency remains quite impressive, though power use overall is up just a bit from the regular 4070. It's still hard to swallow a graphics card that costs more than an entire Xbox Series X or PlayStation 5, a GPU that can easily cost just as much as the entire rest of your PC tower combined. The 4070 Super also needs to contend with AMD's Radeon 7800 XT, which is often as fast or faster, has an additional 4GB of memory, and can often be had for as little as $510. But there's still no card that's offering quite this combination of efficiency and performance. In 2024's GPU market, where stubbornly higher prices are an enduring hangover of the 2020–2021 GPU shortage, a significant performance bump for a not-even-year-old card that's available for the same price counts as a win. Meet the RTX 4070 Super The 4070 Super Founders Edition is black with black highlights, removing the silver highlights from the regular 4070. Andrew Cunningham The standard array of DisplayPort and HDMI outputs. Andrew Cunningham The thing that makes the 4070 Super super is the addition of 1,280 extra CUDA cores, a nearly 22 percent increase from the regular 4070. That's more of an increase in computing power than the 4080 Super or 4070 Ti Super saw, both in absolute core counts and expressed as a percentage. That doesn't always mean a 22 percent bump in-game performance, but it's a nice uplift, and it's a whole lot closer to the AD104 GPU die's maximum core count of 7,680. RTX 4090 RTX 4080 RTX 4080 Super RTX 4070 Ti RTX 4070 Ti Super RTX 4070 RTX 4070 Super CUDA Cores 16,384 9,728 10,240 7,680 8,448 5,888 7,168 Boost Clock 2,520 MHz 2,505 MHz 2,550 MHz 2,610 MHz 2,610 MHz 2,475 MHz 2,475 MHz Memory Bus Width 384-bit 256-bit 256-bit 192-bit 256-bit 192-bit 192-bit Memory Clock 1,313 MHz 1,400 MHz 1,437 MHz 1,313 MHz 1,313 MHz 1,313 MHz 1,313 MHz Memory size 24GB GDDR6X 16GB GDDR6X 16GB GDDR6X 12GB GDDR6X 16GB GDDR6X 12GB GDDR6X 12GB GDDR6X TGP 450 W 320 W 320 W 285 W 285 W 200 W 220 W The only other specs that really change are cache, which is boosted from 36MB in the 4070 to 48MB in the 4070 Super, and the total graphics power (TGP) of the card, which can now reach a maximum of 220 W. It's a small enough difference that Nvidia's required power supply wattage remains the same at 650 W. The 4070 Super retains the same 192-bit memory interface and 12GB bank of memory as the regular 4070, and memory and GPU core clocks are all the same, too. For this review, we tested the Founders Edition version of the card, which conforms to Nvidia's default specs and uses the 12VHPWR connector, but cards from Nvidia's partners can come with mild overclocks and 8-pin power connectors. The included 12VHPWR adapter only needs two 8-pin connectors attached, or you can connect it directly to a compatible power supply; ATX 3.0 supplies support 12VHPWR natively, but I recently bought an adapter cable for my gaming PC's modular Cooler Master PSU that plugs directly into the 8-pin connectors on the power supply; you've got more options now than you did a year ago, now that the 40-series has been around for a bit. Nvidia's Founders Edition card is the same size and shape as the old 4070, which means that (unlike the 4080 and 4090 Founders Edition cards and a whole bunch of triple-fan, triple-slot third-party cards) it ought to fit in just about anything without much hassle, including most mini ITX builds. The only real design difference is that the silver highlights have been swapped out for black ones, giving the card a more imposing "this is my final form" aesthetic. Similarly, we'd expect most of Nvidia's partners to continue using their 4070 card designs without many changes. Performance and power Gaming testbed CPU AMD Ryzen 7 7800X3D (provided by AMD) Motherboard ASRock X670E Taichi (provided by AMD) RAM 32GB (2x16GB) G.Skill Trident Z5 Neo (provided by AMD), running at DDR5-6000 SSD Western Digital Black SN850 1TB (provided by Western Digital) Power supply EVGA Supernova 850 P6 (provided by EVGA) CPU cooler 280 mm Corsair iCure H115i Elite Capellix AIO Case Lian Li O11 Air Mini OS Windows 11 22H2 with Core Isolation on, Memory Integrity off Drivers Nvidia RTX 4070 Super and 3080 Ti: Pre-release driver 546.52 (12/18/2023) Other Nvidia cards: 536.99 (8/8/2023) AMD RX 7700 XT/7800 XT: Pre-release driver 23.20.01.05 Other AMD cards: Adrenalin 23.8.1 If the RTX 4070 went toe to toe with the RTX 3080, then the 4070 Super tries to compete with higher-end 3000-series cards like the RTX 3080 Ti (according to our testing) and RTX 3090 (according to Nvidia's—we don't have one on hand). Both of these 30-series cards perform pretty similarly in many games, though the 3080 Ti might be a better comparison point for the 4070 Super since it also has 12GB of RAM instead of 24GB. And the 4070 Super provides that performance while using roughly two-thirds as much power, according to our tests. Per usual, we've run a variety of DirectX 11 and DirectX 12 games on these GPUs, focusing on 1440p and 4K resolutions since that's what the card is suited for. We've also included charts for games with heavy ray-tracing effects like Cyberpunk 2077 and Returnal, and also some charts measuring performance with DLSS/FSR and DLSS Frame Generation enabled. See our RTX 4090 review for an explainer about DLSS FG. 3DMark benchmarking numbers and 1440p tests. Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Compared to the regular 4070, adding 22 percent more CUDA cores without changing most of the card's other specs generally makes the 4070 Super about 15 percent faster in most games, though you may rarely hit that 22 percent number (we did once, in a 4K Cyberpunk 2077 test with ray tracing on). You'll also see many examples throughout our test suite, though, where the performance gains are smaller, in the 8–10 percent range. Generally, though, it's a nice upgrade. Comparing the 4070 Super and 3080 Ti in more depth, we see quite a few games where the cards run equally fast. At 1440p, in fact, the 4070 Super usually pulls just about even with the 3080 Ti, or it falls short by just a handful of frames. Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Bump the resolution up to 4K, though, and the 3080 Ti's edge gets more consistent. The 4070 Super still keeps up most of the time, and it has the (admittedly imperfect) extra benefit of DLSS Frame Generation to lean on. But the 3080 Ti has a 384-bit memory interface, and the 4070 Super's 48MB of cache can't quite close that gap at higher resolutions. On the topic of DLSS FG, the 4070 Super is kind of an ideal card for it. To get good results, your GPU generally needs to be capable of a reasonably good base frame rate, something that's much more true of the 4070 Super at 1440p and 4K than it is of the 4060 Ti or 4060. You'll still see some weird visual artifacts (the rainstorm in Returnal that blinks in and out of existence remains a good, easily reproducible example), but it's worth turning it on in more demanding games like Cyberpunk 2077 to see whether the benefit of the higher frame rate is offset by visual glitches or increased input lag. Andrew Cunningham Andrew Cunningham Andrew Cunningham Andrew Cunningham Finally, let's compare the 4070 Super to AMD's best competitor in this category, the Radeon RX 7800 XT. Generally available for just a hair over its $500 MSRP, the 7800 XT can usually match or beat the 4070 Super's performance in games without ray-tracing effects enabled. It's not always true, but it usually is. Per usual, though, AMD comes out worse in games with heavy ray-tracing effects, falling well behind the 4070 Super (or even the regular 4070) in Cyberpunk, Returnal, Hitman III, and the 3DMark ray-tracing benchmarks. Andrew Cunningham Like the 4070, the 4070 Super remains an incredibly power-efficient GPU. Its absolute power consumption in games is between 10 and 15 percent higher than the regular 4070 in our tests, roughly commensurate with the performance boost you can usually expect. The 7800XT uses an additional 30 W or so running the same games at the same settings (as ever, take software-reported power consumption comparisons between manufacturers with a grain of salt, but these numbers are in line with what both Nvidia and AMD list in their spec sheets). Conclusions The GeForce RTX 4070 Super is a pretty straightforward win for most GPU buyers. It's the same price as the 4070 and has better performance, and it's a $600 card that trades blows with cards like the 3080 Ti and 3090 ($1,199 and $1,499 at launch, respectively). Comparing it to the 3080 Ti, you can still see games where the 4070 Super's 192-bit memory interface seems to be holding it back, particularly at 4K. But the architectural updates and extra processing power generally help the 4070 Super keep up. It simplifies matters for most GPU buyers that we are mostly past the transitional phase of this graphics generation where you wouldn't just have to factor in what the competition was offering but also all the last-gen GeForce and Radeon cards that were still sitting on shelves at reasonable prices. At this point, it's not worth considering the leftover, open-box stock of the 6800 XT or 3080 Ti. One problem with the 4070 Super is that it still leaves a fair-sized hole in Nvidia's GPU lineup, where none of the options are particularly cost-effective. There are now three cards between the 4070 Super and the $299 RTX 4060: the 8GB 4060 Ti at $399, the 16GB 4060 Ti at $449, and the regular RTX 4070 at $549. The 4060 Ti is generally between 10 and 20 percent faster than the 4060 for 33 percent more money, and the 4070 Super is usually between 12 and 22 percent faster than the regular 4070 for about 9 percent more money. Either card is one decent promotional discount away from being a solid deal, but at their regular prices, you should save a bit of money with the regular 4060, spend a bit more to get the 4070 Super, or consider one of AMD's offerings. AMD remains pretty competitive here, offering its typical bargain: pay a little less for a card with as-good-or-better rasterized performance but higher power use and generally so-so ray tracing performance. If you want high-frame-rate 1440p or reasonably solid 60 fps 4K, the 7800 XT is still a viable option, and it costs closer to $500 than $600. The power consumption, better ray-tracing performance, and DLSS support are all strong arguments in favor of the 4070 Super or even the regular 4070, but if you don't need or care about any of those things, go ahead with the AMD card and save some money. Great GPU for high-fps 1440p gaming, with 60 fps 4K well within reach for many games Excellent power efficiency RTX 3080 Ti-esque performance for half the price Nvidia's retains its performance edge in ray-traced games Does rely on DLSS to hit 60 fps when playing at 4K or at 1440p with heavy ray-tracing effects on Relatively narrow 192-bit memory bus means it can't always beat the old 3080 Ti, especially at 4K If you don't care about ray tracing, DLSS, or lower power consumption, the $500-ish Radeon RX 7800 XT may be a better value $600 is still a price increase from older xx70 GPUs Source

On January 8, the big three in the PC hardware space, AMD, Intel, and Nvidia, all revealed new products. AMD unveiled its Ryzen 8000G series desktop APUs, its first on the DDR5 platform, as well as much less exciting 7600 XT, a 16GB variant of the RX 7600. Meanwhile, Intel released the rest of its desktop 14th Gen lineup and these aren't as power-hungry as the K SKUs (A new driver is out as well). It also released Raptor Lake mobile U and HX parts. The latter, Intel has claimed is often faster than AMD's fastest mobile gaming APU, the one with the 3D V-cache. Out of the three, Nvidia definitely had the most impressive product reveal as it launched three new top-of-the-line desktop GPUs and also knocked down the price by a significant amount. The trio of new RTX 4000 Super GPUs: the RTX 4080 Super, the 4070 Ti Super, and the 4070 Super boast higher shaders as well as more VRAM in some cases. Nvidia also disclosed the specifications of the GPUs, and the company, inadvertently seemed to have under-specced its RTX 4070 Super (spotted by Wccftech). After stating in its press release that the card features 36MB of level 2 (L2) cache, Nvidia has since quietly updated the spec sheet of the GPU and it now reads 48 MB of L2, which is an increase of 33%. Nvidia took a page from AMD's book and included this large pool of cache this time with its Ada Lovelace design. This cache pool helps compensate for memory bandwidth while also decreasing cache misses. Before After You can view the rectified L2 cache specification in the before/after image above. The before image shows the 36MB and the after image shows 48MB (the full AD104 GPU die). While we aren't exactly sure, the lower cache amount may have been a result of a miscommunication between the technical team and the marketing team as a specification upgrade in such short notice is not possible. Source

As January starts to settle in, tech enthusiasts are probably aware that it is CES month. As such, companies too are starting to tease upcoming products and some of them have already made their announcements public. Nvidia is, so far, in the former camp, and a few days ago, the company published its first such teaser in the form of a new Facebook cover photo to start building the hype for its upcoming 4000 Super series GPUs. The Super series is not common in Nvidia's product launch cycle and it only started with the 2000 series when the company announced a refresh of its 20-series and these new cards came equipped with more CUDA cores and sometimes more VRAM. Nvidia is doing something similar to that this time too as the new cards, at least some of them, will have more VRAM as one of the major complaints from most of the tech press has been about the provided memory buffer, which is simply not enough for the price points Nvidia launched the cards at. Since the cards' announcements are only a couple of days away, Nvidia and its AIB partners are now testing out the GPUs. As such, the RTX 4070 Super's performance has now leaked as X user/bot Benchleaks spotted two entries of the GPU on Geekbench. The 4070 Super was tested on both OpenCL and CUDA, and the GPU managed to score 195,384 and 219,237 points respectively. Although this is not exactly the 3D rasterization or ray tracing performance of the GPU, the compute scores here typically tend to align quite nicely with how the 3D output of the graphics card is. OpenCL CUDA Courtesy of Wccftech, we can see how the new 4070 Super goes up against the RTX 4080, the 4070 Ti, and the vanilla 4070. As expected from the purported specifications of the GPUs, it is there sandwiched between the 4070 and the 4070 Ti. Sadly, we can't compare how the 4070 Super will fare against AMD cards with these numbers. For that, we will have to wait either for more benchmarks leak or the review NDA uplift. Source: Geekbench (link1, link2) via Benchleaks (X1, X2) Source

New rumor suggest that Nvidia could make two new variants of the RTX 4070 graphics card. One a cut-down version and another a RTX 4070 SUPER. While the RTX 4090 gets praised everywhere due to its unmatched performance, the other graphics cards in the RTX 4000 series don’t get that many compliments. This is because while RTX 4090 is massively faster than its predecessor, that’s not a case for other graphics cards like RTX 4080, RTX 4070 & RTX 4060 series, which have rather unattractive improvement in the performance. If that is not enough, then it’s likely that we will not see the next-gen Nvidia RTX 5000 series graphics card anytime soon, as they are stated to release only two years later. So it’s highly likely that Nvidia wants to refresh its RTX 4000 series with improved specifications in the graphics card series. This could be done releasing the refreshed versions in between the release of the RTX 4000 and RTX 5000 series. Something it has previously done in both the GTX 1600 and RTX 2000 series. A few days ago, a well known leaker MEGAsizeGPU revealed that Nvidia could be working on a RTX 4080 Ti or RTX 4080 SUPER graphics card, with improved specs by using a more powerful GPU and it’s expected to release next year, in 2024. Yesterday, VideoCardz reported that Nvidia could be working on three new graphics card in the RTX 4000 series, the RTX 4080 SUPER, RTX 4070 Ti SUPER and RTX 4070 SUPER. With release date being unknown. Looks like we have gotten some information about one of them, the Nvidia GeForce RTX 4070 SUPER. RTX 4070 SUPER Specs Leaked Today, MEGAsizeGPU on Twitter (via VideoCardz) has shared information about the possible specs of two new graphics cards in the RTX 4070 series. The tweet mentions that Nvidia might be making two new graphics cards. The first one is going to be the RTX 4070 D6, which would be based on the same specs as the current RTX 4070, but will use a slower GDDR6 VRAM. For the record, RTX 4070 uses faster and better GDDR6X memory. This could mean that RTX 4070 could be available for a cheaper price, but with slower VRAM. More importantly, MEGAsizeGPU reveals that Nvidia is working on another graphics card upgrade, the RTX 4070 SUPER. As per the tweet, the RTX 4070 SUPER will be based on the AD103 GPU, which is the same GPU used in the RTX 4080 graphics card. Additionally, the RTX 4070 SUPER would come with a 256-bit memory bus and 16GB VRAM. For the record, the normal RTX 4070 graphics card uses AD104 GPU and comes with just 12GB of VRAM using 192-bit memory bus. The price and the CUDA cores count of the RTX 4070 SUPER isn’t revealed by MEGAsizeGPU as of now. But expect it to be higher than the normal RTX 4070 and expect an increase in the performance too. This is welcomed because the RTX 4070 reviews weren’t that great and wasn’t selling much. So much so that Nvidia had even paused its production. So if Nvidia does decide to improve the RTX 4070 and other graphics cards in the series, while keeping it at an attractive price point, then it would be a great thing for the users. Source