A Chinese manufacturer is now selling a modified GeForce RTX 5090 with 96 GB of memory, three times the 32 GB on the retail card. Shenzhen Suqiao Intelligent Technology, an OEM/ODM with over a decade of experience building graphics cards, motherboards, and servers, listed the card on Alibaba for $3,888. That's roughly 35% less than a standard RTX 5090 currently costs in the US, and a fraction of what NVIDIA charges for its RTX PRO 6000 Blackwell. The latter uses the same GB202 die and now starts at a whopping $16,000.

Suqiao likely built this with a custom PCB running clamshell memory, packing GDDR7 modules on both sides of the board to double capacity, the same approach NVIDIA itself uses on the Pro 6000. The GPU die may come from harvested RTX 5090 units, reballed onto the new board and running on leaked or modified firmware, a method Chinese workshops already used to push the RTX 4090 to 48 GB VRAM. The listing does raise some doubts, though, since it mentions GDDR6X memory at 14 Gbps, which doesn't line up with a genuine 96 GB Blackwell build.

A Brazilian modder going by fmklab has built a working RTX 3070 Ti with 16 GB of memory, a configuration NVIDIA reportedly planned years ago but never brought to market. Instead of bolting bigger memory chips onto a retail 3070 Ti board, Fabio, aka fmklab, started with an ASUS TUF RTX 3070 PCB, which already runs on regular GDDR6 rather than the GDDR6X found on the real 3070 Ti. The RTX 3070 GPU was pulled and replaced with an actual 3070 Ti chip, and the original Samsung HC14 (8 Gb) memory modules were swapped for HC16 (16 Gb) units to double capacity. GPU-Z still reported 8 GB after the swap, so fmklab used NVIDIA's BIOS Reader tool to work out which memory straps needed rewiring. A few tweaks later, the full 16 GB buffer was showing up.

Once that was sorted, MATS testing came back clean, and the card ran a local LLM using nearly the full 16 GB pool. Moreover, Call of Duty: Warzone paired with a Ryzen 7 9800X3D was pulling over 13 GB with no signs of instability. NVIDIA had reportedly explored a 16 GB RTX 3070 Ti as far back as 2020, and the idea resurfaced through leaked engineering samples before NVIDIA reportedly shelved it. Fabio's Superposition score landed a bit below typical retail 3070 Ti runs, likely because GDDR6 is slower than GDDR6X, a fair tradeoff for doubling the memory this way. The base RTX 3070 has seen similar VRAM mods before, but this looks like the first working recreation of this specific 16 GB 3070 Ti variant.

Rowhammer attacks exploit vulnerabilities in DRAM hardware by causing targeted bit-flips, allowing attackers to bypass memory isolation and gain control over a device. These attacks were initially focused on CPUs and their associated DDR memory, such as DDR4. However, recent research indicates that NVIDIA GPUs are also vulnerable due to the fragile nature of the GDDR6 memory they use, which directly compromises the CPU host. Two independent research teams have discovered ways to exploit this decades-old memory vulnerability against modern graphics hardware, with trouble extending beyond the GPU itself. The "GDDRHammer" and "GeForge" groups have each developed functional exploits that use Rowhammer bit-flips in NVIDIA GPUs with GDDR6 memory to gain complete control over the host CPU's memory. This attack can perform bit-flips on some NVIDIA GPU models, ranging from the "Ampere" to "Ada Lovelace" families of cards.

An attacker who succeeds in this can read and write anything stored in the machine's main memory. Both teams have also introduced new Rowhammer techniques specifically designed for GPU architecture, achieving a significantly higher rate of bit-flips on GDDR6 memory than previous methods. The critical step in both exploit chains involves targeting the GPU's memory allocator, using controlled bit-flips to corrupt the GPU's page tables. Once these page tables are compromised, the attacker gains arbitrary read and write access to CPU memory, breaking down the security boundary between the graphics subsystem and the rest of the machine. The end result is a full system compromise, as the attacker can manipulate memory at will and gain root access, achieving total control without interacting with privileged software paths. The affected GPUs include the GeForce RTX 3060, which experienced 1,171 bit-flips, and the RTX 6000 "Ada" GPU, which saw 202 bit-flips from the attack.

NVIDIA's "Ampere" generation of GeForce RTX graphics cards continues to attract buyers, albeit mostly across used markets. Certain discerning gaming GPU enthusiasts often consider newer "Ada Lovelace" and "Blackwell" architectures to be too expensive. Seninno, an AI PC hardware enthusiast, attempted to strike the right balance between price and past-gen performance. His mid-November video report focused on an unusual custom GeForce RTX 3080 card. Under normal circumstances, owners of RTX 30-series sub-flagships could play around with 10 GB of GDDR6X VRAM. Another unusual 20 GB example has come to light—courtesy of Seninno's search for a bargain. Historically, a handful of variants—mostly made by MSI—have turned up via used product channels in China. Seninno's video report delves into his acquiring of an ASUS ROG Strix GeForce RTX 3080 20 GB card.

Weeks after performing an extensive teardown, Seninno's discovery gained wider press coverage—yesterday's Uniko's Hardware article weighed in on debates over the card's origins. Could it be an engineering sample, an OEM release, or a previous owner's modification project? The YouTuber stated: "I found it (for sale in a) Chinese flea market, I've never seen this model, and I can't found any info online, so I decided to try it, but the condition wasn't great, or I would've kept it." An inspection of printed material pointed to a fairly official looking "STRIX-RTX3080-20G" code. A "10DE 2206 -1043 87AA" device ID was detected by GPU-Z, with a "20480 MB" memory size. Seninno was not impressed with the card's scuffed condition, both externally and internally—it was likely hooked up to a crypto mining rig in the past—and aims to return it to an unnamed vendor. Upon examining the deshrouded card, he reckoned that the memory modules were intact and original, but there was evidence of thermal paste reapplications and screws being removed, and not replaced.

A Russian technician and modder known as VIK-on has posted a hands-on video showing how Chinese workshops are converting GeForce RTX 4090 gaming cards into AI-focused accelerators by doubling their memory. Using an MSI RTX 4090 Suprim as a donor, VIK-on removed the AD102 GPU package and the card's 12 2 GB GDDR6X chips, then placed those parts onto a custom clamshell PCB that accepts memory on both sides. The builders pair the board with a blower-style cooler, designed for multi-GPU racks, and load modified firmware so that the system recognizes the larger capacity. The method mixes factory-level board work with careful firmware changes and advanced soldering. You can purchase a brand-new custom PCB that includes clamshell memory support and has most passive components pre-installed. You can see it in the third picture below, the upper PCB is the newly bought one, the bottom one is the original. However, the GPU and memory must be reflowed or hand-soldered, which requires expensive tools and expertise.

Firmware mods, some of which trace back to techniques used for salvaging cards with faulty memory channels, are adapted to enable the GPU to accept the new layout. Shops often harvest GDDR6X chips from defective cards or purchase modules at a low cost, which reduces expenses, but the conversion process is still not trivial. Published price notes indicate that the clamshell kit costs roughly $140 in China, and the total cost of the parts for a DIY upgrade is near $430 before shipping.
Converting large numbers of consumer cards for AI reduces the supply for gamers, and stripped or modified boards can resurface on the second-hand market, surprising buyers. Additionally, modified cards may not meet reliability standards or all warranty rules. They can even malfunction under sustained workloads. Therefore, it is better to skip going after a customized GPU to save yourself the hassle of compatibility issues; however, the point that it can be done remains.

A previously unseen AMD graphics-card prototype has unexpectedly surfaced, not in a retailer's warehouse but on a Korean second-hand forum. This week, a Quasarzone user known as FP32 shared high-resolution photographs of what appears to be the cooling assembly for a never-released "RX 7950 XTX," allegedly sourced via China's Xianyu marketplace. At a staggering 34 cm in length and 5.5 cm in thickness, the triple-slot shroud outstrips the standard RX 7900 XTX reference cooler by roughly 5 cm in length and adds an entire slot of girth. The front houses three recessed eight-pin power connectors, suggesting board power well over 450 W, a figure more commonly associated with NVIDIA's RTX 4090 rather than an AMD Radeon, while a dense copper baseplate feeds multiple heatpipes that terminate in three striking red-tipped fins stamped "RDNA 3." Despite its superficial resemblance to AMD's shipping designs, the prototype's mounting points and LED cut-outs align with no known PCB, and the customary rear I/O bracket is entirely missing.

Attempts to trace serial labels in public component databases yielded no matches, reinforcing the theory that it was a one-off engineering sample. No PCB accompanied the cooler, preventing any live demonstrations or performance benchmarks. There have long been whispers about an AMD halo card, whether an RX 7950 XTX or even a dual-GPU RX 7990 XTX, to bridge the performance gap between NVIDIA's 80-class and 90-class offerings. The emergence of this monolithic cooler implies that AMD once considered a fully unlocked Navi 31 silicon pushed to higher clock speeds or paired with faster GDDR6/GDDR6X memory. The project may have been shelved due to soaring production costs, unwieldy thermal envelopes, or a strategic decision to focus on energy efficiency and value-oriented designs. Whatever the reason for its cancellation, this orphaned cooler now stands as a relic of AMD's road not taken, offering a glimpse into the company's highest ambitions before it chose a less aggressive path for its RDNA 3 lineup.

GALAX introduced the GeForce RTX 4080 SUPER 3X, a custom-design RTX 4080 SUPER graphics card that appears to be its most affordable product based on the GPU. It's positioned a notch below the SG, ST, EX Gamer, and HOF series. The card features a sporty, multi-layered plastic cooler shroud, and a triple-fan setup. The cooling solution is no more than 3 slots thick, and features what looks like two aluminium fin stacks on either ends of a set of heat pipes.

The PCB only covers two-thirds the length of the card, and so all of the airflow from the third fan, and some of it from the second one, go through the heatsink, and out back from a cutout in the backplate. The fans lack any kind of lighting, the only lighting is an illuminated GeForce RTX logo on top. The card lacks any factory overclocking, but GALAX includes a software-based "1-click OC" mode. The RTX 4080 SUPER maxes out the "AD103" silicon, enabling at 10,240 CUDA cores, 80 RT cores, 320 Tensor cores, 112 ROPs, and 320 TMUs present on the silicon, besides the full 64 MB L2 cache available. The GPU is wired to 16 GB of 23 Gbps GDDR6X memory across a 256-bit wide memory interface, yielding 736 GB/s of memory bandwidth.

NVIDIA quietly released a variant of the GeForce RTX 4070 featuring slower 20 Gbps GDDR6 memory, replacing the 21 Gbps GDDR6X that the original RTX 4070 comes with. Wccftech has access to a GALAX branded RTX 4070 GDDR6, and put it through benchmarks focused on comparing it to a regular RTX 4070. Memory type and speed are the only changes in specs, the core-configuration isn't changed, nor is the GPU clock speed. Wccftech's testing shows that the RTX 4070 GDDR6 measures within 0-1% slower than the RTX 4070 (GDDR6X) at 1080p and 1440p resolutions; while the difference between the two is about 2% at 4K Ultra HD.

Wccftech's test-bed is comprehensive, with 27 game tests, each across 3 resolutions; and 7 synthetic tests. The synthetic tests are mainly part of the 3DMark test suite, including Speed Way, Fire Strike, Time Spy, Port Royal, and their presets. Here, the RTX 4070 GDDR6 is nearly identical in performance, with a 0-0.2% delta with the RTX 4070 GDDR6X. In the game tests, performance varies by resolution. 1080p has 0-1% performance delta, with the only noteworthy outliers being "Metro Exodus" (extreme preset), where the RTX 4070 GDDR6 loses 4.2%, and "Alan Wake 2," where it loses 2.3%.

NVIDIA GeForce board partners are preparing a silent launch of a variant of the GeForce RTX 4070 with slower 20 Gbps GDDR6 memory in place of the 21 Gbps GDDR6X that's standard to the RTX 4070, which results in a 5% reduction in memory bandwidth. It turns out that other specs, such as GPU clocks or core-configuration aren't changed to compensate for the reduced memory bandwidth. ASUS is among the first board partners with an RTX 4060 GDDR6 card, the ASUS DUAL RTX 4070 GDDR6, which was briefly listed on Newegg for $569, before it went out of stock. This is reported by VideoCardz as being the same price as the regular ASUS DUAL RTX 4070 with GDDR6X.

ASUS isn't the only NVIDIA board partner with an RTX 4070 GDDR6, Wccftech spotted a GALAX branded card that comes with the model string "RTX 4070 D6 1-click OC." Its retail box features a large specs-sheet on the front face that clearly mentions GDDR6 as the memory type. NVIDIA's move to re-spec the RTX 4070 with 20 Gbps GDDR6 was originally seen as a move to reduce its costs, letting the card be sold closer to the $500-mark. It remains to be seen if real-world prices settle down below those of the original RTX 4070 cards.

ASUS became the first NVIDIA add-in card partner to debut a GeForce RTX 4070 graphics card with GDDR6 memory, away from its original memory type of GDDR6X. The RTX 4070 originally comes with 12 GB of 21 Gbps GDDR6X memory, yielding 504 GB/s of memory bandwidth. This ASUS card (which we hope is part of a larger refresh by NVIDIA to reduce costs of the RTX 4070), uses GDDR6 of a yet-unspecified speed. You do have 21 Gbps GDDR6 (non-X) in the market, but we imagine these to be expensive, which leaves NVIDIA with the 20 Gbps GDDR6 chip that's more readily available (which AMD uses in its RDNA 3 graphics cards), and 18 Gbps GDDR6 that the company itself uses in cards such as the RTX 4060 Ti. Any reduction in memory bandwidth would have to be compensated with increase in GPU clocks, but we don't see that happening here—the regular variant of the ASUS DUAL RTX 4070 GDDR6 EVO comes with the reference 2475 MHz maximum boost speed, while the DUAL OC variant only slightly cranks this up to 2520 MHz.

GIGABYTE introduced the GeForce RTX 4070 Ti SUPER WindForce MAX graphics card. It is characterized by an oversized air cooling solution that gives it some large dimensions of 33.1 cm length, 5.55 cm thickness (3 slots), but more importantly, a height of 13.6 cm, which could pose a challenge for those with mid-tower cases, given the tight cable bending restrictions of the 12VHPWR connector. GIGABYTE found a novel solution to this problem. The power connector is located pointing toward the tail end of the card, where the PCB terminates. The PCB is only two-thirds the length of the card, and so the power cable can be routed in without any bends.

The WindForce cooling solution features a trio of 100 mm fans that ventilate a large aluminium fin-stack heatsink with nine copper heatpipes, and a direct-touch base. GIGABYTE has given the card a factory overclock of 2655 MHz GPU clocks, compared to 2610 MHz reference, while leaving the memory untouched at 21 Gbps. The card offers dual-BIOS, with the default BIOS enabling these clock speeds, and the Silent BIOS lowering them to reference speeds, while quietening the cooler. Based on the 5 nm AD103 silicon, the RTX 4070 Ti SUPER is endowed with 8,448 CUDA cores, 66 RT cores, 264 Tensor cores, 96 ROPs, and 264 TMUs. The GPU gets 16 GB of 21 Gbps GDDR6X memory across a 256-bit wide memory interface. The company didn't reveal pricing.

PC enthusiasts who crave powerful graphics in compact systems have a new option from Zephyr. The Chinese manufacturer has launched the world's first ITX-sized GeForce RTX 4070 graphics card. Dubbed the Sakura Blizzard, this GPU packs NVIDIA's AD104-250 chip and 12 GB of GDDR6X memory into a footprint of just 172 x 123 x 42 mm. While slightly taller than standard PCIe brackets, the two-slot cooler should fit most Mini-ITX cases. The card's cute pink shroud and solitary cooling fan give it a unique aesthetic. But don't let the pink looks fool you - Zephyr claims this compact powerhouse can keep the GPU and memory up to 10°C cooler than some dual-fan RTX 4070 designs, which needs to be taken with a grain of salt. Thermal testing videos show the fan spinning at 2,400 RPM to maintain GPU temperatures around 73°C under load in a 25°C room. Meanwhile, synthetic benchmarks reportedly demonstrate no performance compromises versus full-sized RTX 4070 implementations.

Zephyr's initial production run has already sold out in China. However, a second batch is slated for mid-July availability to meet the apparent higher demand for small form factor RTX 40-series GPUs. The launch comes just weeks after NVIDIA unveiled new "SFF-ready" design guidelines at Computex 2024. As the power-hungry RTX 40 lineup hit the market, many voiced concerns over the cards' ever-growing dimensions. NVIDIA's renewed SFF PC focus signal options like the Sakura Blizzard could become more common. For space-constrained enthusiasts, having top-tier GPU muscle in a properly-cooled Mini-ITX card is a big win. Zephyr's ITX-sized RTX 4070 shows powerful things can come in small packages, and we hope more manufacturers follow this philosophy.

The race is on for memory manufacturers to bring the next generation GDDR7 graphics memory into mass production. While rivals Samsung and Micron are aiming to have GDDR7 chips available in Q4 of 2024, South Korean semiconductor giant SK Hynix revealed at Computex 2024 that it won't kick off mass production until the first quarter of 2025. GDDR7 is the upcoming JEDEC standard for high-performance graphics memory, succeeding the current GDDR6 and GDDR6X specifications. The new tech promises significantly increased bandwidth and capacities to feed the appetites of next-wave GPUs and AI accelerators. At its Computex booth, SK Hynix showed off engineering samples of its forthcoming GDDR7 chips, with plans for both 16 Gb and 24 Gb densities.

The company is targeting blazing-fast 40 Gbps data transfer rates with its GDDR7 offerings, outpacing the 32 Gbps rates its competitors are starting with on 16 Gb parts. If realized, higher speeds could give SK Hynix an edge, at least initially. While trailing a quarter or two behind Micron and Samsung isn't ideal, SK Hynix claims having working samples now validates its design and allows partners to begin testing and qualification. Mass production timing for standardized memories also doesn't necessarily indicate a company is "late" - it simply means another vendor secured an earlier production window with a specific customer. The GDDR7 transition is critical for SK Hynix and others, given the insatiable demand for high-bandwidth memory to power AI, graphics, and other data-intensive workloads. Hitting its stated Q1 2025 mass production target could ensure SK Hynix doesn't fall too far behind in the high-stakes GDDR7 race, with faster and higher-density chips to potentially follow shortly after volume ramp.

Micron is getting ready to set sail with the new GDDR7 memory, and being NVIDIA's long-term partner, both are gaining big with the GeForce RTX 4000 Series. A few days ago, Micron announced its imminent launch of GDDR7 memory, remember that in March JEDEC published the GDDR7 memory standard. The interesting part is that an even more important piece of information was provided in one of Micron's slide-deck, the one that shows GDDR7 gaming performance. Officially Micron says that GDDR7 can be 3.1x faster than GDDR6 and 1.5x faster than GDDR6X. However, the slide in question provides some normalized FPS performance for ray tracing, and rasterization under game testing conditions. Until now, everything is clear, what comes next is our speculation and assumptions, based on previous knowledge so please, take it with a grain of salt. Will it be a hard guess, looking over those graphs, to tell that NVIDIA's RTX 5090 offers up to a 42% improvement over the RTX 4090 in rasterization, and up to 48% in ray tracing?

NVIDIA is getting closer to releasing its next-gen GPU with the RTX 5090 graphics card. Micron is in a strong partnership with NVIDIA and it's expected that its next gen graphics cards will use Micron GDDR7 memory. Now, we know that, at least for Ray Tracing, one real-world benchmark is Cyberpunk 2077, and it could be also the case here, from the graphs it looks very much like Ray Tracing was tested with Path Tracing. For Rasterization performance another hard guess is that perhaps it's either Starfield or Assassin's Creed Mirage. Presumably, tests were made using also AMD RX 7900XTX, and NVIDIA RTX 4090 to get GDDR6 reference. Time will tell if we made the right prediction.

At Computex 2024, Colorful has introduced a new PC series of premium componentry for games, called iGame Ultra Series. Carrying a premium white aesthetics, the Ultra Series serves as clean and elegant component for PC builds. At the heart of the Ultra Series is the GeForce RTX 4080 SUPER graphics card. The card has a triple-fan cooling solution with vapor chamber technology that keeps temperatures under control. Powering this GPU is the iGame Z790D5 Ultra motherboard. Supporting the latest Intel 14th Gen CPUs and up to 192 GB of DDR5-7800 RAM. An array of ultra-fast PCIe 5.0 and 4.0 M.2 slots allow multiple high-speed SSDs for ridiculous storage performance. Keeping all of these elite components running is the 850 W iGame P850G Ultra W 80 Plus Gold power supply. Its modular design with dedicated PCIe 5.0 connectors ensures clean cable management and stable power delivery. For cooling the CPU, there is the iGame LQ360 Ultra W all-in-one liquid cooler, which sports a massive 360 mm radiator and a trio of high-static pressure fans. All of these premium components are housed in the iGame C23A Ultra W full-tower chassis. With a spacious dual-compartment layout, front USB-C port, and preinstalled vertical GPU bracket, it offers both incredible expandability and sleek aesthetics.

The graphics card modders from Brazil, TecLab and Paulo Gomes are back with yet another audacious, competitive VGA mod. The duo has already built a reputation for video memory replacement mods that have significantly altered the performance profile of graphics cards, and their latest feat sees an NVIDIA GeForce RTX 4070 Ti SUPER graphics card get a bit of a memory upgrade. Out of the box, the RTX 4070 Ti SUPER comes with 16 GB of 21 Gbps GDDR6X memory across the 256-bit wide memory interface of the "AD103" GPU. The two tech-tubers each performed memory chip replacement mods, which in combination with GPU overclocking, resulted in memory speeds ranging between 24 Gbps and 26 Gbps. The big story here is that with faster memory, the RTX 4070 Ti SUPER beats the RTX 4080, despite 13% fewer shaders and other key components.

A stock GeForce RTX 4080 graphics card scores 8525 points in the Unigine Superposition 8K benchmark, and a stock-speed Manli RTX 4070 Ti SUPER does 7212 points, on account of fewer shaders (66 SM vs. 76 SM) and slower memory (21 Gbps vs. 22.4 Gbps). With a 24 Gbps memory speed mod, and GPU overclocking, Paulo Gomes achieved 8870 points in the test for the RTX 4070 Ti SUPER, beating the RTX 4080, and overcoming the shader deficit. Meanwhile, over at TecLab, their Galax-branded RTX 4070 Ti SUPER yields 7028 points at stock speeds; and the team pulled off an epic 26 Gbps memory speed (chip replacement + overclocking), which when combined with some GPU overclocking, yielded a staggering 9133 points, which would inch close to what an RTX 4080 SUPER could produce with its 23 Gbps memory and maxed out "AD103" ASIC with all 80 SM on deck.

Palit Microsystems Ltd, a leading graphics card manufacturer, is thrilled to unveil the White Series for the GeForce RTX 4070 and RTX 4060 Ti graphics cards. The White series offers a perfect blend of elegant looks and uncompromised power. The models feature a sleek white color scheme across both the shroud and backplate, easily integrating into any light-colored PC build. The design caters to the growing desire for a clean and sophisticated look for your PC.

The twin 95 mm fans and optimized backplate vents ensure effective thermal performance. Equipped with 0dB technology, the cards minimize the noise during light loads like multimedia applications and browsing. During demanding tasks, fans automatically adjust to optimal cooling and performance. The Palit GeForce RTX 4070 and RTX 4060 Ti White Series cater to both gamers and creators. Gamers will benefit from the exceptional performance boost, allowing them to experience high frame rates and smooth gameplay. Creators will appreciate the powerful processing capabilities for tasks like video editing and high-resolution design work. Additionally, the sleek white aesthetics perfectly complement a modern creative workspace.

NVIDIA is confirmed to implement the GDDR7 memory standard with the top three GPU ASICs powering the next-generation "Blackwell" GeForce RTX 50-series, Tweaktown reports, citing XpeaGPU. By this, we mean the top three physical silicon types from which NVIDIA will carve out the majority of its SKUs. This would include the GB202, the GB203, and GB205; which will power successors to everything from the current RTX 4070 to the RTX 4090. NVIDIA is expected to build these chips on the TSMC 4N foundry node.

There will be certain GPU ASIC types in the "Blackwell" generation that will stick to older memory standards such as GDDR6 or even the GDDR6X. These would be successors to the current AD106 and AD107 ASICs, powering SKUs such as the RTX 4060 Ti, and below. NVIDIA co-developed the GDDR6X standard with Micron Technology, which is the chip's exclusive supplier to NVIDIA. GDDR6X scales up to 23 Gbps and 16 Gbit, which means NVIDIA can avail plenty of performance for the lower-end of its product stack using GDDR6X; especially considering that its GDDR7 implementation will only run at 28 Gbps, despite chips being available in the market for 32 Gbps, or even 36 Gbps. Even if NVIDIA chooses the regular GDDR6 standard for its entry-mainstream chips, the tech scales up to 20 Gbps.

The first round of NVIDIA GeForce RTX 50-series "Blackwell" graphics cards that implement GDDR7 memory are rumored to come with a memory speed of 28 Gbps, according to kopite7kimi, a reliable source with NVIDIA leaks. This is despite the fact that the first GDDR7 memory chips will be capable of 32 Gbps speeds. NVIDIA will also stick with 16 Gbit densities for the GDDR7 memory chips, which means memory sizes could remain largely unchanged for the next generation; with the 28 Gbps GDDR7 memory chips providing 55% higher bandwidth over 18 Gbps GDDR6 and 33% higher bandwidth than 21 Gbps GDDR6X. It remains to be seen what memory bus widths NVIDIA chooses for its individual SKUs.

NVIDIA's decision to use 28 Gbps as its memory speeds has some precedent in recent history. The company's first GPUs to implement GDDR6, the RTX 20-series "Turing," opted for 14 Gbps speeds despite 16 Gbps GDDR6 chips being available. 28 Gbps is exactly double that speed. Future generations of GeForce RTX GPUs, or even refreshes within the RTX 50-series could see NVIDIA opt for higher memory speeds such as 32 Gbps. When the standard debuts, companies like Samsung even plan to put up fast 36 Gbps chips. Besides a generational doubling in speeds, GDDR7 is more energy-efficient as it operates at lower voltages than GDDR6. It also uses a more advanced PAM3 physical layer signaling compared to NRZ for JEDEC-standard GDDR6.

Some of the first gaming GPUs that implement the next-generation GDDR7 memory standard, will stick to 16 Gbit memory chip densities (2 GB), according to kopite7kimi, a reliable source with NVIDIA GeForce leaks. 16 Gbit is what is standard for the current RTX 40-series graphics cards, which ensures that a GPU with 256-bit memory bus gets 16 GB of video memory; the ones with 192-bit get 12 GB; and the ones with 128-bit get 8 GB. The flagship RTX 4090 uses twelve of these chips over its 384-bit memory bus for 24 GB.

Kopite7kimi's leak could have a different connotation, that much like the RTX 30-series "Ampere" and RTX 40-series "Ada," NVIDIA might not use JEDEC-standard GDDR7 on all product segments, and might co-engineer an exclusive standard with a DRAM company with memory bus signaling and power management technologies most optimal to its graphics architecture. It co-developed the GDDR6X with Micron Technology to do exactly this. GDDR7 comes with data-rates as high as 32 Gbps, which will be the top speed for the first round of GDDR7 chips that come out toward the end of 2024, heading into 2025. The second round of GDDR7 chips slated for late-2025 going into 2026, could go as fast as 36 Gbps. This is similar to how the first GDDR6 chips were 14-16 Gbps, and the next round did 18-20 Gbps.

KFA2, the brand owned by Galax for specific markets in Europe, announced its GeForce RTX 4070 SUPER EX Gamer graphics card in two new trims, Pink, and White. The regular RTX 4070 SUPER EX Gamer the company launched in January comes in black. The Pink trim involves a bubblegum pink shade taking over the cooler shroud and metal backplate. Ditto with the white trim. The PCB underneath both cards remains black, but since the cooling solution dwarfs it, it most remains out of sight when installed.

Both cards feature frosted acrylic fan impellers, with addressable RGB LEDs located in the fan hubs. Besides the illuminated fans, the GeForce RTX logo on top of the card comes with RGB LED illumination. Both cards come with a mobile app-based control for the lighting. This app talks to a service installed in your PC over the Internet, which interacts with the graphics card. Much like the regular black RTX 4070 SUPER EX Gamer, the two cards stick with NVIDIA-reference clock speeds of 2475 MHz boost for the GPU, and 21 Gbps (GDDR6X-effective) for the memory. KFA2 is pricing both the Pink and White variants at 679€ including taxes, a 20€ premium over the regular black EX Gamer card.

ZOTAC USA Inc., a leading manufacturer of innovative gaming hardware solutions, is thrilled to announce the market launch of three new MEK HERO pre-built Gaming PCs: the MEK HERO A7647S, A7647STI, and A7748S built using the latest ZOTAC GAMING GeForce RTX 4070 SUPER Twin Edge OC, ZOTAC GAMING GeForce RTX 4070 Ti SUPER Trinity OC White Edition, and ZOTAC GAMING GeForce RTX 4080 SUPER Trinity Black Edition Graphic Cards. These are teamed up with AMD Ryzen 5 7600 and Ryzen 7 7700X processors, offering stunning 4K graphics for unprecedented gaming performance with AI-accelerated DLSS 3.5 and Real-Time Ray Tracing, bringing gamers an immersive gaming experience. This new series of MEK HERO offers 3 years of warranty coverage, specifically for the included ZOTAC GAMING GeForce RTX 40 SUPER Series Graphics Card component.

The MEK HERO A7647S, A7647STI, and A7748S Gaming PCs feature 16 GB DDR5 system memory and NVMe M.2 SSD storage, elevating PC gaming and the visual content creation process to new heights. Each MEK HERO Gaming PC undergoes meticulous assembly and hand-testing in the United States, ensuring the highest build quality. Crafted for optimal performance, durability, and an unmatched experience, MEK HERO Gaming PCs with GeForce RTX 40 SUPER Series Graphics Cards are tailored to meet the demands of hardcore gamers and content creators alike.

PNY announced today the arrival of the new VERTO GeForce RTX 4080 SUPER 16GB, RTX 4070 Ti SUPER 16GB, and RTX 4070 SUPER 12GB graphics cards to its lineup of NVIDIA GeForce RTX GPUs. The latest generation of RTX, GeForce RTX SUPER 40-series graphics cards are blazingly fast, offering gamers and creators an unparalleled boost in performance, neural rendering, and many more cutting-edge platform features. They are fueled by the revolutionary NVIDIA Ada Lovelace architecture; a major advancement in GPU technology which empowers accelerated content production techniques, amazing AI capabilities, and hyper-realistic gaming experiences.

The new GeForce RTX SUPER GPUs are the ultimate way to experience AI on PCs. Specialized AI Tensor Cores deliver up to 836 AI TOPS to deliver transformative capabilities for AI in gaming, creating and everyday productivity. PC gamers demand the very best in visual quality, and AI-powered NVIDIA Deep Learning Super Sampling (DLSS) Super Resolution, Frame Generation and Ray Reconstruction combine with ray tracing to offer stunning worlds. With DLSS, seven out of eight pixels can be AI-generated, accelerating full ray tracing by up to 4x with better image quality.

NVIDIA today gave its GeForce RTX 40-series "Ada" a midlife refresh targeting the higher end of its product stack, with the new GeForce RTX 4070 SUPER, GeForce RTX 4070 Ti SUPER, and the GeForce RTX 4080 SUPER. The new RTX 4080 SUPER replaces the current RTX 4080, which will gradually be phased out of the market. The new RTX 4070 Ti SUPER does the same to the current RTX 4070 Ti. The RTX 4070 SUPER, however, will coexist with the current RTX 4070, albeit at a slight price premium. The RTX 4070 SUPER and RTX 4070 Ti SUPER are both being recommended by NVIDIA for maxed out 1440p gaming with full ray tracing; while the RTX 4080 SUPER is for those who want to max out gameplay at 4K with full ray tracing. The RTX 4070 SUPER and RTX 4070 Ti SUPER should still very much be capable of 4K gaming and more than acceptable frame rates, especially given the latest DLSS 3 Frame Generation and its proliferation among new AAA titles.

NVIDIA is giving the three new graphics card SKUs a staggered launch spread across January 2024. The RTX 4070 SUPER should be available to purchase on January 17, at a starting price of $599, which was the original MSRP of the RTX 4070. After this launch, the RTX 4070 slides down a bit to $549 while remaining in the product stack. Things get interesting higher up the stack. The RTX 4070 Ti SUPER, which goes on sale on January 24, is priced at $799, while the current RTX 4070 Ti is being retired from the product stack. The remaining RTX 4070 Ti cards should be up at slightly discounted prices.

Here are some of the first pictures of the ASUS GeForce RTX 4070 SUPER Dual OC, the company's close-to-MSRP custom-design implementation of the upcoming RTX 4070 SUPER, which is expected to be announced on January 8, with reviews and retail availability a week later. The card very closely resembles the design of the RTX 4070 Dual OC, but with one major difference—the single 8-pin PCIe power connector makes way for a 16-pin 12VHPWR. Considering that the ASUS Dual OC series tends to come with a nominal factory OC at power limits matching NVIDIA reference, this is the first sign that the RTX 4070 SUPER in general might have typical graphics power (TGP) above what a single 8-pin could fulfill, and so we've given a 12VHPWR, just like every RTX 4070 Ti. The cards will include an NVIDIA-designed adapter that converts two 8-pin PCIe to a 12VHPWR, with its signal pins set to tell the graphics card that it can deliver 300 W of continuous power.

The GeForce RTX 4070 SUPER is based on the same AD104 silicon as the RTX 4070 and RTX 4070 Ti, with its ASIC code rumored to be "AD104-350." The SKU allegedly enables 56 out of 60 streaming multiprocessors (SM) present on the silicon, giving it 7,168 out of 7,680 CUDA cores. This is a big increase from the 5,888 CUDA cores (46 SM) that the vanilla RTX 4070 is configured with. The memory subsystem is expected to be unchanged from the RTX 4070 and RTX 4070 Ti—12 GB of 21 Gbps GDDR6X across a 192-bit memory interface; leaving NVIDIA with one possible lever, the ROP count. While the RTX 4070 Ti has 80 ROPs, the RTX 4070 has 64. It remains to be seen how many the RTX 4070 SUPER gets. Its rumored TGP of 225 W is behind the switch to 12VHPWR connectors.