News Posts matching #RTX 30-series

Return to Keyword Browsing

GeForce Unlocked: Modders Boost RTX 20, RTX 30, and RTX 40 Compute Performance by Up to 2x

NVIDIA's GeForce RTX 30-series and RTX 40-series graphics cards are based on the same dies as their professional-grade GPUs. This means the compute potential of consumer-oriented RTX GPUs is restricted by NVIDIA's firmware stack, but modders have managed to unlock the instruction issuing rate equivalent to their professional GPU counterparts. Starting with the GeForce RTX 30-series "Ampere," a researcher developed an unlock that effectively doubles the instruction issuing speed of FP/BF16 FMA into FP32 operations on GA102, GA103, GA104, and GA106 cards. As a result, cards like the GeForce RTX 3090 Ti experience up to 80% faster matrix multiplication in BF16, which is crucial for running nearly all neural networks. The unlock supports GA102, GA103, GA104, and GA106-based GeForce RTX graphics cards.

Next, we have the GeForce RTX 40-series "Ada Lovelace" with another project that has demonstrated a massive 2× speedup in FP16 and BF16 tensor throughput with FP32 accumulation. When the GeForce RTX 4090 was tested with FP16 tensor throughput and matrix accumulation, the GPU achieved 165.6 TeraFLOPS, while the unlock pushed that number to 331.6 TeraFLOPS. This is intriguing as these computational performance numbers are typically reserved for the higher-end professional segment like the NVIDIA RTX 6000 "Ada Lovelace." It seems NVIDIA artificially limited the AD102 silicon's performance in compute-sensitive matrix multiplication and accumulation, affecting the performance of running large language models. For this unlock, AD102, AD103, and AD106 are officially unlocked.

New Tools Let Users Edit and Load Modified VBIOS on GeForce RTX GPUs Without Flashing

A developer who goes by Panchovix has released three tools that let NVIDIA GeForce RTX 50, RTX 40, and RTX 30-series owners run a modified VBIOS without writing anything to the card's flash chip. The first is a VBIOS editor for Windows and Linux that can change power limits, voltage ceilings, current (OCP) limits, fan channels, and the core and memory offset ranges. Edited images keep NVIDIA's original signature, which no longer matches, so they can't be flashed with nvflash. Patched tools did break NVIDIA's BIOS signature lock for crossflashing back in 2023, but this approach skips flashing altogether. On Windows, the loader stores the edited image in the GPU's driver registry key, where the driver picks it up as an override every time the device starts, and it only needs a quick GPU restart. The Linux loader instead patches NVIDIA's open kernel modules so the GSP firmware starts with the custom image, with Fedora 43 being the only distro tested on real hardware so far.

The screenshots show how this works in practice. On what appears to be an RTX 5090 Founders Edition ROM, the board power ceiling went from 600 W to 700 W and the NVVDD and MSVDD voltage ceilings from 1,110 mV to 1,150 mV. An RTX 4090 ROM had its power ceiling raised from 666 W to 750 W and its core offset range doubled to +2,000 MHz, and nvidia-smi later confirmed that card's power limit going from 550 W to 671 W. Apparently, an RTX 5090 reached 800 W, an RTX 5080 went from 360 W to 500 W, and an RTX 5060 Ti from 185 W to 300 W in the developer's own testing. Laptop RTX 40 GPUs can't be edited yet, and memory timings can't be changed since the card reads them directly from the flash chip.

Modders Bring NVIDIA's Faster DLSS Frame Generation Preset C to RTX 30-Series GPUs

NVIDIA's new DLSS Frame Generation Preset C has only been out for a few days, and modders have already brought it to RTX 30 cards. The latest version of dashdogy's Universal MFG Unlocker (v1.5.0, formerly RTX 40 MFG Unlock) adds Preset C support on both RTX 40 and RTX 30 GPUs, including 3x to 6x multi-frame generation (MFG). Officially, Preset C is a beta "performance mode" for RTX 40, RTX 50, and RTX Spark systems, available through the NVIDIA app's experimental features. It swaps one of Frame Generation's two AI models, the one handling optical flow, for a lighter algorithm, and NVIDIA claims up to 35% higher performance in exchange for a small hit to image quality.

According to the release notes, RTX 30 support is now considered stable, with smoother frame pacing at 3x and above, though it only works in DirectX 12 games. However, the newest DLSS Frame Generation build that carries Preset C has only been validated offline on RTX 30, and the preset was confirmed in-game on an RTX 40 card in The Witcher 3. No RTX 30-series performance numbers have been shared yet, so the gains on Ampere are anyone's guess. The update also adds XeSS Multi Frame Generation from 2x to 6x on NVIDIA, AMD, and Intel GPUs through an OptiScaler plugin, and fixes a launch crash in Gears of War: E-Day. This effort follows recent mods that got DLSS Multi Frame Generation running on Ampere and also brought Smooth Motion to the same cards.

NVIDIA Smooth Motion Reaches GeForce RTX 30-Series "Ampere" GPUs Through Mods

Modders have gotten NVIDIA's Smooth Motion frame generation running on GeForce RTX 30 series cards, VideoCardz reports. Smooth Motion is NVIDIA's driver-level frame generation feature, similar in spirit to AMD's AFMF, and NVIDIA only exposes it officially on RTX 40-series and newer through the NVIDIA App. Ampere owners have been left out, at least officially. Two unrelated projects are apparently working to close that gap.

The first comes from developer pipotoufikxyz-lgtm, who expanded the earlier dlssg_for_sm86 project, itself an adaptation of NVIDIA's DLSS-G runtime for the SM86 architecture that RTX 30 GPUs use. The update adds Smooth Motion and Vulkan support, and it reportedly went live on September 19, tested on an RTX 3070. The second, NVSmooth30, comes from a developer going by ItsAdeline and takes a different route: rather than building on the existing DLSS-G package, it's a clean-room compatibility layer aimed directly at NVIDIA's NvPresent64.dll Smooth Motion path. Currently, NVSmooth30 has been tested on an RTX 3050 laptop GPU, and separately on an RTX 3070.

Developer Ports DLSS 5 to Intel Integrated Graphics With Help From AI Agents

A developer going by Uzbekunknown on GitHub has gotten NVIDIA's DLSS 5 Neural Rendering running on the Arc 140V integrated graphics of an Intel Lunar Lake system, as Tom's Hardware reports. The project, called dlss-nr-on-intel, is an independent reimplementation of the 71-block network that runs on Intel's XMX units through Vulkan. It has apparently been shown working in Tekken 7, Dead or Alive 5 Last Round, and Mortal Kombat 1. For now it only runs on Linux, and users must supply their own copy of NVIDIA's DLL. As far as performance goes, the developer says Tekken 7 ran at around 10 FPS at 640x360 resolution, and the frametimes are absolutely abysmal too. The author calls it a research port, not a finished product, so photo mode use looks like the only realistic scenario right now.

Perhaps even more interesting is how this port came to be. Apparently, the code was written almost entirely by AI. The developer credits Claude Opus 5 and OpenAI's Astra, saying they themselves only supplied the hardware, the binary, and the direction. They also state they cannot defend the code line by line, and that a driver bug one of the agents hallucinated shaped three development phases. This mod effort follows DLSS 5 reaching RTX 30-series and AMD RDNA 4 cards within weeks of the leak, as well as the technology reaching web browsers. I think whether this is clever or a step too far depends on how much you trust code nobody fully reviewed. It's a fine technical demo, but I don't think we really needed DLSS 5 on Intel integrated graphics.

Modders Unlock DLSS Multi Frame Generation For RTX 30-Series "Ampere" GPUs

After porting DLSS 5 to every recent NVIDIA generation, the modding community now seems to be turning its focus to Frame Generation. A new mod called DLSSG SM86 lets GeForce RTX 30-series owners run DLSS Multi Frame Generation in 2X and 4X modes, a feature NVIDIA officially reserves for RTX 50-series "Blackwell" GPUs. Unlike earlier workarounds that swap in AMD's FSR Frame Generation, this project keeps NVIDIA's own DLSS models intact. It works by creating a proxy backend for the RTX 30-series' SM86 architecture that runs alongside the game and redirects frame generation calls to a bundled DLSSG 310.1 runtime, without touching the game's own files.

The developer's documentation lists the RTX 3080 Ti as the tested card, running driver 591.86 under Windows with D3D12. Reported gains include Black Myth: Wukong climbing from 50 FPS to 150 FPS at 4X Frame Gen, and Cyberpunk 2077 with path tracing going from 35 FPS to 100 FPS. That said, the developer is upfront that formal frametime, latency, and long-duration stability testing hasn't been done yet, so real-world smoothness don't fully match those raw numbers. A YouTube video by fogoticus also shows the feature being tested on an RTX 3080, with Cyberpunk 2077 jumping from around 42 FPS to 74 FPS in 2X mode, then to 120 FPS in 4X mode.

New DLSS 5 Swapper Tool Brings Neural Rendering to Games NVIDIA Never Supported

A new tool called DLSS 5 Swapper, published on GitHub by developer rakanki911, automates installing DLSS 5's Neural Rendering into games, including ones that never officially got support for it. Users can drag and drop a game into the app, or let it scan the drive for installed titles on its own. For games that already ship with DLSS, the swap is simple. For everything else, the tool relies on DLSS5-Feeder, a community mod that emulates the DLSS calls required to trigger Neural Rendering, paired with ReShade for the depth buffer and motion data that DLSS needs to work. That combination reportedly runs across titles from DirectX 9 through DirectX 12, and even on a handful of emulators like PCSX2, Dolphin, and Xenia.

Perhaps even more interestingly, the app also extends DLSS 5 support to RTX 20-series and RTX 30-series graphics cards, sidestepping NVIDIA's RTX 50-series requirement. This tool builds on the same leaked DLSS 5 files that surfaced in NBA 2K27's pre-launch build, files modders have spent weeks getting to run on older GPUs and non-DLSS games. DLSS 5 Swapper mostly bundles that scattered work into a single app; however, since it's unofficial and has no ties to NVIDIA, it's important to verify the download source before installing anything. Given how far these mods have already spread, this tool is likely to become the most popular way for people to try Neural Rendering, at least until NVIDIA offers something similar itself.

DLSS 5 Officially Goes Live September 3 With NBA 2K27, Exclusive to RTX 50-series "Blackwell" GPUs

NVIDIA has confirmed that DLSS 5 will officially launch on September 3 alongside NBA 2K27, and it will remain exclusive to GeForce RTX 50-series GPUs. When asked directly at a Gamescom technical briefing whether older cards would ever receive support, NVIDIA gave a flat answer: "It's RTX 50-series only." The model runs on Blackwell's fifth-generation Tensor cores, and there's no fallback path for RTX 40-series or anything older. That confirmation lines up with what modders had already found by digging through the leaked DLSS 5 DLL that surfaced through an early NBA 2K27 build last week, which contained executable code gated specifically to Blackwell hardware.

According to NVIDIA, DLSS 5 is its largest and most complex model to date, and the priority has been optimizing it specifically for RTX 50 hardware rather than stretching it across older Tensor core generations. However, that didn't stop modders from finding workarounds for RTX 40-series, 30-series, and even 20-series GPUs. After the DLSS 5 DLL leaked through an early NBA 2K27 build, a modder patched around the CUDA binaries and got Neural Rendering running on an RTX 4090, with performance dropping by roughly 36 to 39 percent in Control. That's still a significant hit, but nowhere near the collapse seen on RTX 30-series cards, which lack native FP8 support and saw frame rates fall off a cliff in the same tests. Given that the RTX 40-series already has native FP8 acceleration built in, official support for it seems like it could've been on the table, even as a lower-tier experience

Leaked DLSS 5 is Being Tested on RTX 20-series "Turing" GPUs, Now Runs on Emulators and Old DirectX Titles

The leaked DLSS 5 DLL keeps finding new homes, and support has now spread to RTX 20-series "Turing" GPUs. That architecture is three whole generations behind the NVIDIA-recommended "Blackwell," and modders are showing no signs of stopping either. Developer ShortFuse has released new DLSS-NR builds for older RTX hardware, including one described as a "speed boost" build for RTX 20-series. The results are still a mixed bag, however. Some games run, while others crash or fail to initialize. Red Dead Redemption reportedly worked, while Red Dead Redemption 2 did not, and Hogwarts Legacy got stuck in a standby state instead of activating the feature. GTA 5 ran for about ten seconds before the display signal dropped out entirely.

Moreover, X user @DystopianSuns says the community now has DLSS 5 running on RTX 20 and RTX 30 GPUs using an FP16 implementation. This is a change from the FP8 approach that previously tanked performance on Ampere cards. FP16 should suit Turing and Ampere better, since neither generation accelerates FP8 the way Blackwell does. With this approach, users reported respectable performance in many titles on the RenoDX Discord using RTX 30-series cards. The same X user apparently got Manhunt running through the PS2 emulator PCSX2 with DLSS 5 enabled, and other testers report the mod working in older titles built on DirectX 9 and DirectX 11, well outside the APIs NVIDIA has confirmed for the feature.

Leaked DLSS 5 Reaches RTX 30 Series "Ampere" GPUs, But Performance Falls Apart

DLSS 5 leaked through an early-access build of NBA 2K27 last week, and since then, the modding community has been hard at work patching it into all sorts of different games. Modders on the RenoDX Discord have now pushed it into RTX 30-series "Ampere" GPUs, despite this generation having no official support for DLSS 5. The results, however, are less than stellar. On the RenoDX Discord, a user with an RTX 3070 laptop GPU running Kingdom Come: Deliverance II saw render latency jump from roughly 29 ms to over 3,300 ms with Neural Rendering enabled, resulting in well under 1 FPS.

Over on X, user @Flourek got it running on an RTX 3080, and it fared little better. It too, however, dropped from around 130 FPS to just 4 FPS in Deep Rock Galactic, while RTX 3050 and RTX 3060 Ti cards reportedly hovered near 1 FPS across Control, FiveM, and Kingdom Come: Deliverance II. Just a couple of days ago, modders got DLSS 5 running on RTX 40-series "Ada Lovelace" cards, and performance held up noticeably better. As we noted in our DLSS-NR deep dive, Ampere lacks native FP8 support, which the leaked DLSS 5 model apparently relies on, so the steep drop isn't surprising. Officially, DLSS 5 only supports RTX 50-series "Blackwell" GPUs, and there's no word about whether NVIDIA even plans to add support for RTX 40-series cards, let alone older ones.

NVIDIA Releases GeForce 572.16 WHQL Game Ready Drivers

NVIDIA today released the latest version of its GeForce Game Ready drivers. Version 572.16 WHQL comes with support for the latest NVIDIA GeForce RTX 5080 and RTX 5090 "Blackwell" graphics cards, along with all the software framework needed for DLSS 4 and Multi Frame Generation. DLSS 4 and its updated Transformer-based models for super resolution, ray reconstruction, and frame generation are even available to older generations of GeForce RTX GPUs depending on their hardware capability. The updated super resolution model is available across the GeForce RTX series, the updated ray reconstruction to RTX 40-series and RTX 30-series; and the updated frame generation model to the RTX 40-series. Multi frame generation is exclusive to the RTX 50-series.

A bug that caused G-SYNC compatible monitors to flicker when framerate drops below 60 FPS has been fixed. Micro-stutters noticed with "Indiana Jones and the Great Circle" with V-sync disabled, has been fixed. Stability has been improved for Ubisoft games that use the Snowdrop game engine. The list of games with DLSS 4 multi frame generation support can be found in the table below.
DOWNLOAD: NVIDIA GeForce 572.16 WHQL

NVIDIA Plans RTX 3050 A with Ada Lovelace AD106 Silicon

NVIDIA may be working on a new RTX 3050 A laptop GPU using an AD106 (Ada Lovelace) die, moving away from the Ampere chips used in other RTX 30-series GPUs. While not officially announced, the GPU is included in NVIDIA's latest driver release and the PCI ID database as GeForce RTX 3050 A Laptop GPU. The AD106 die choice is notable, as it has more transistors and CUDA cores than the GA107 in current RTX 3050s and the AD107 in RTX 4050 laptops. The AD106, used in RTX 4060 Ti desktop and RTX 4070 laptop GPUs, boasts 22.9 billion transistors and 4,608 CUDA cores, compared to GA107's 8.7 billion transistors and 2,560 CUDA cores, and AD107's 18.9 billion transistors and 3,072 CUDA cores.

While this could potentially improve performance, it's likely that NVIDIA will use a cut-down version of the AD106 chip for the RTX 3050 A. The exact specifications and features, such as support for DLSS 3, remain unknown. The use of TSMC's 4N node in AD106, instead of Samsung's 8N node used in Ampere, could potentially improve power efficiency and battery life. The performance of the RTX 3050 A compared to existing RTX 3050 and RTX 4050 laptops remains to be seen, however, the RTX 3050 A will likely perform similarly to existing Ampere-based parts as NVIDIA tends to use similar names for comparable performance levels. It's unclear if NVIDIA will bring this GPU to market, but adding new SKUs late in a product's lifespan isn't unprecedented.

NVIDIA RTX 20-series and GTX 16-series "Turing" GPUs Get Resizable BAR Support Through NVStrapsReBAR Mod

February saw community mods bring resizable BAR support to several older platforms; and now we come across a mod that brings it to some older GPUs. The NVStrapsReBAR mod by terminatorul, which is forked out of the ReBarUEFI mod by xCurio, brings resizable BAR support to NVIDIA GeForce RTX 20-series and GTX 16-series GPUs based on the "Turing" graphics architecture. This mod is intended for power users, and can potentially brick your motherboard. NVIDIA officially implemented resizable BAR support since its RTX 30-series "Ampere" GPUs in response to AMD's Radeon RX 6000 RDNA 2 GPUs implementing the tech under the marketing name Smart Access Memory. While AMD would go on to retroactively enable the tech for even the older RX 5000 series RDNA GPUs, NVIDIA didn't do so for "Turing."

NVStrapsReBAR is a motherboard UEFI firmware mod. It modifies the way your system firmware negotiates BAR size with the GPU on boot. There are only two ways to go about modding a platform to enable resizable BAR on an unsupported platform—by modding the motherboard firmware, or the video BIOS. Signature checks by security processors in NVIDIA GPUs make the video BIOS modding route impossible for most users; thankfully motherboard firmware modding isn't as difficult. There is an extensive documentation by the author to go about using this mod. The author has tested the mod to work with "Turing" GPUs, however, it doesn't work with older NVIDIA GPUs, including "Pascal." Resizable BAR enables the CPU (software) to see video memory as a single contiguously addressable block, rather than through 256 MB apertures.

NVIDIA Celebrates 500 Games & Apps with DLSS and RTX Technologies

NVIDIA today announced the important milestone of 500 games and apps that take advantage of NVIDIA RTX, the transformative set of gaming graphics technologies, that among many other things, mainstreamed real-time ray tracing in the consumer gaming space, and debuted the most profound gaming technology of recent times—DLSS, or performance uplifts through high-quality upscaling technologies. The company got to this milestone over a 5-year period, with RTX seeing the light of the day in August 2018. NVIDIA RTX is the combined feature-set of real time ray tracing, including NVIDIA-specific enhancements; and DLSS.

Although it started out as an upscaling-based performance enhancement that leverages AI, DLSS encompasses a whole suite of technologies aimed at enhancing performance at minimal quality loss, and in some cases even enhances the image quality over native resolution. This includes super resolution, or the classic DLSS and DLSS 2 feature set; DLSS 3 Frame Generation, which nearly doubles frame-rates by generating entire alternate frames entirely using AI, without involving the graphics rendering machinery; and DLSS 3.5 Ray Reconstruction, which attempts to vastly improve the fidelity of ray traced elements in upscaled scenarios.

NVIDIA GeForce 537.42 WHQL Released with Support for DLSS 3.5 and Cyberpunk 2077 Phantom Liberty

NVIDIA today released the GeForce 537.42 WHQL Game Ready drivers. These introduce support for the new DLSS 3.5 Ray Reconstruction technology, which will work on all GeForce RTX GPUs, including RTX 20-series "Turing," RTX 30-series "Ampere," and the latest RTX 40-series "Ada." Ray Reconstruction is a data-driven DNN-based ray tracing denoiser that produces superior results than hand-tuned legacy denoisers. In particular, ray traced elements such as shadows and reflections in games that use DLSS 2 (super resolution), should look more natural. The driver also comes with optimization for "Cyberpunk 2077: Phantom Liberty," besides "Warhaven," "Witchfire," and "Party Animals." The drivers also fix a couple of performance issues with Octane Renderer.

DOWNLOAD: NVIDIA GeForce 537.42 WHQL

Be sure to check out our Performance and Image Comparison review of DLSS 3.5 Ray Reconstruction later today, and our technical overview of how it works.

Galax Reportedly Preparing GeForce RTX GPU Price Cuts

A recent report published by BoardChannels points to Galax possibly implementing a broad set of price cuts across its range of NVIDIA GeForce RTX 40 and 30-series custom graphics card models. Insider information originating from sources within a pool of NVIDIA and AMD board partners suggests that Galax could be shaving off up to 1000 RMB (around $140) from certain Ampere and Ada Lovelace products - effective later this month in its native Hong Kong market as well as mainland China.

The article posits that GeForce RTX 4080 cards could end up becoming 1000 RMB cheaper, and popular RTX 3060 models receiving cuts of around 250 RMB (≈$35). Galax is reported to have already offered entry-level desktop GeForce RTX 3050 cards at lower prices in the latter half of May - with 140 RMB (≈$19.50) reductions. The RTX 4070 series is supposedly set to receive a measly discount of around 150 to 200 RMB (≈$21 to $28), which is likely not doing it many favors given slow worldwide uptake since the product range's launch in mid-April. Galax could be making adjustments to fall in line with rivals (in the region) who have already reduced asking prices for NVIDIA gaming hardware.

Age of Wonders 4 Gets a Bunch of Hotfixes on PC

Paradox Interactive states: "To my fellow Godir. First off, we want to thank you for the awesome release! It's been amazing seeing so many of you finally getting your hands on the game. Many of you have been able to enjoy the game as we intended but there have been difficulties along the way. Over the past few days we've been hard at work to address the most common issues that have popped up. The people on the Hotfix branch have been incredibly helpful in tracking down the Graphics Driver issues and we're now confident in rolling out the fixes at large."

For users with the NVIDIA Graphics cards defined below, that have experienced frequent crashes, we have added a Driver Workaround setting. This setting will be automatically enabled for the affected users and should alleviate the most commonly experienced crashes. May your Pantheons Thrive

NVIDIA GeForce Game Ready Drivers 531.79 WHQL Released

NVIDIA has released the latest iteration of its GeForce Game Ready drivers - Version 531.79 WHQL. This latest update provides optimizations for today's launch of Redfall - Arkane's horror adventure shooter features DLSS 3 technology - ensuring the best possible in-game performance on day one. This Game Ready Driver provides DLSS 2 technology support for a number of other launch titles including Showgunners and the upcoming Diablo IV Server Slam (May 12). Owners of RTX 30-series graphics cards have previously reported application crashes in Battlefield 2042, The Last of Us and the Unreal Editor - today's patch notes indicate that these issues have been fixed by the new update. Owners of Asus VG27WQ and Acer XV253Q monitors will be pleased to know that the highest available display refresh rate is now restored within NVIDIA's Control Panel app - the option was reported to have disappeared around version 531.41.

DOWNLOAD: NVIDIA GeForce Game Ready 531.79 WHQL

Component Suppliers Suggest That NVIDIA is Taking a Relaxed Approach with RTX 40-Series Production

Two of NVIDIA's providers of Outsourced Semiconductor Assembly and Test (OSAT) services are of the opinion that Team Green is happy to stay the course with its Ada Lovelace GPU production schedule. The backend providers Siliconware Precision Industries (SPIL) and King Yuan Electronics (KYEC) have not been given any new instructions with regard to shifts (up or down) in component assembly output. It is theorized that NVIDIA is aiming to clear any stock backlogs of graphics card models featuring previous generation architecture - namely the second gen GeForce RTX 30-series, built on Ampere.

The retail demand for the newly released GeForce RTX 4070 cards has been mild, to say the least - with plenty of inventory remaining on the shelves in the States. Critical reception of the midweight GeForce RTX GPU has also been middling - many have advised that budget conscience buyers should potentially look elsewhere. The market for discrete graphics card is in a fairly healthy state at the moment, with major production issues and fractured supply chains becoming lesser concerns for electronics manufacturers. NVIDIA has the advantage of being a market leader, and seems to be quite content with proceedings - but their analysts are very likely keeping an eye on the RTX 4070 sales figures. Its products are out and readily available - no need to change direction too sharply.

NVIDIA pulls RTX 30-series Founders Edition Cards From its Webstore

NVIDIA has removed all its GeForce RTX 30-series Founders Edition graphics cards from its webstore. There are still plenty of RTX 30-series graphics cards available everywhere, including the NVIDIA webstore, as NVIDIA is happily selling custom versions from its AIC partners. This does not mean that you can't find the RTX 30-series Founders Edition graphics cards at other shops, as some might be available from other retailers/e-tailers like Amazon, Best Buy, Newegg, Mindfactory, and others. It is now obvious that NVIDIA is slowly phasing out the Ampere-based graphics cards and that it will focus on GeForce RTX 4090, 4080, and 4070 Ti, as well as the GeForce RTX 4070 that should come soon.

NVIDIA GeForce 526.47 WHQL Game Ready Drivers Released

NVIDIA today released the latest version of its GeForce Game Ready drivers. Version 526.47 WHQL adds support for two new GPUs, namely the GeForce RTX 3060 8 GB, and the RTX 3060 Ti GDDR6X, the two SKUs NVIDIA launched to improve its standing against the RX 6650 XT, RX 6600, and Arc 7-series. Among the game optimizations with this release are for "Sackboy: A Big Adventure," "Victoria 3," "WRC Generations," and DLSS 3 frame-generation support in F1 22. Among the handful issues fixed with this release are game map corruption in "Cyberpunk 2077," a crash and reboot issue noticed with the Dell XPS 9560; lower performance noticed in Minecraft Java Edition; the 165 Hz refresh-rate option not being available with Samsung Odyssey Ark monitors; GeForce Experience selecting the wrong display-head with Shadowplay; and certain online video artifacting noticed with NVIDIA Image Scaling enabled.

DOWNLOAD: NVIDIA GeForce 526.47 WHQL

NVIDIA Partners Quietly Launch GeForce RTX 3060 with 8GB (128-bit) Memory

NVIDIA's add-in board partners today began quietly launching the GeForce RTX 3060 8 GB, a variant of the RTX 3060 with a third of its memory size and memory bus-width sawed off. The RTX 3060, NVIDIA's best-selling desktop graphics SKU from the RTX 30-series "Ampere," originally launched with 12 GB of GDDR6 memory across a 192-bit wide memory bus, which at its reference speed of 15 Gbps (GDDR6-effective), makes 360 GB/s of memory bandwidth. The new variant comes with 8 GB of GDDR6 memory across a narrower 128-bit memory interface, with the same 15 Gbps data-rate, which works out to 240 GB/s memory bandwidth.

Besides memory size, bus-width, and bandwidth; NVIDIA hasn't tinkered with the core-configuration with the RTX 3060 8 GB. It still comes with 3,584 CUDA cores across 28 SM, which work out to 112 Tensor cores, 28 RT cores, 112 TMUs, and 48 ROPs. The GPU's base frequency is set at 1320 MHz, and boost frequency at 1777 MHz—same as the original RTX 3060. Even the typical graphics power is unchanged, at 170 W. The new 8 GB variant doesn't replace the original, but is being positioned a notch below it, possibly to compete against the likes of the Radeon RX 6600 (non-XT), and perhaps even the Arc A750.

NVIDIA Removes Hashrate Limiter for RTX 30-series LHR GPUs in the Latest Driver

In a last-ditch effort to clear inventory of its GeForce RTX 30-series "Ampere" graphics cards, NVIDIA has reportedly removed the hashrate limiter in the latest GeForce 522.25 drivers, without mentioning it anywhere in the driver's release notes. A Redditor and GeForce RTX 3080 Ti owner by the username "Timbers007" rested that their card, which launched exclusively as LHR-enabled graphics cards (with no RTX 3080 Ti cards without LHR in circulation); is now achieving double its usual hashrates when benchmarked with ethminer. It's able to put out 112 MH/s, a hashrate only possible with mining software that circumvents the LHR limiter, such as NiceHash or NBMiner.

In the thick of the graphics card shortage in 2020-21, as crypto currency miners were buying up inventory of gaming graphics cards through sophisticated retail bots; NVIDIA attempted to sour the milk for miners by introducing LHR (lite hashrate), a supposedly hardware-level limitation that cripples the mining performance of the GPU. This failed to improve things as NVIDIA accidentally released drivers without the hashrate limiter early on and redacted them, but not before they spread among miners. It was only a titanic crash in crypto-currency values, and the recent Ethereum merge that killed GPU-accelerated mining, which arrested demand, bringing RTX 30-series GPUs to prices more acceptable to gamers, as miners began flooding the market with their used GPUs at much lower prices. Will this improve sales of the RTX 30-series? Unlikely. Miners with RTX 30-series LHR graphics cards who already had the hacks to circumvent the limiter, are dumping their cards.

The GeForce RTX 3090 SUPER That Never Was, Pictured

This is probably the only picture of a GeForce RTX 3090 SUPER Founders Edition graphics card. NVIDIA allegedly decided against this branding, in favor of the RTX 3090 Ti, to designate its maxed-out GA102-based graphics card. With no other "SUPER" SKUs in the RTX 30-series and plenty of "Ti," the company probably thought it wasn't worth the trouble to leave the odd SUPER SKU sticking out at the top. The picture surfaced on the NGA Forums, showing the card with prominent "RTX 3090 SUPER" branding etched along the top frame of the cooler. This card has the same device ID as the RTX 3090 Ti, so GPU-Z detects it as such. It also has identical specs to an RTX 3090 Ti Founders Edition.

NVIDIA CEO Confirms RTX 40-series "Ada" Reveal in September, Launch Aimed at Not Cannibalizing "Ampere"

NVIDIA CEO Jen Hsun Huang, in his Q2 Fiscal-2023 Results call confirmed that the company's next-generation GeForce RTX 40-series graphics cards could be revealed at GTC Fall 2022, to be held next month in September. NVIDIA's launch of the RTX 40-series "Ada" will be unlike those of the previous few generations, despite retaining a "top-down" launch cycle (of launching high-end SKUs first). CEO Jen Hsun says that the first products will be "layered on top" of the current-generation "Ampere" products, so they don't cannibalize the sales of current-generation products.

"Ampere is the most popular GPU we've ever created. It is in the top 15 most popular gaming GPUs on Steam. And it remains the best GPUs in the world, and it will be very successful for some time. However, we do have exciting new next-generation coming and it's going to be layered on top of that. And so, we've taken—we've done two things. We've reduced sell-in to let channel inventory correct and we've implemented programs with our partners to price position the products in the channel in preparation for our next generation," said CEO Jen Hsun Huang. This could mean that the RTX 40-series could see a ramp-up to the various mainstream market segments, and gain volumes from them, only in 2023. The remainder of 2022 could see a high-end debut of the RTX 40-series, selling alongside attractively priced RTX 30-series cards.
Update 09:48 UTC: NVIDIA states that the CEO Keynote for GTC is scheduled for September 20.
Return to Keyword Browsing
Nov 22nd, 2026 11:33 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts