News Posts matching #GDDR7

Return to Keyword Browsing

"GPUBreach" is a Rowhammer Attack for GDDR6-Based NVIDIA GPUs That Bypasses IOMMU

Late last week, we reported on a new series of rowhammer bit-flip attacks targeting GDDR6-based NVIDIA GPUs. Most of these attacks can be mitigated by enabling IOMMU through the BIOS, which restricts the memory regions the GPU can access on the host system, thereby closing the primary attack path. However, researchers from the University of Toronto have introduced "GPUBreach," which can bypass IOMMU and enable CPU-side privilege escalation, unlike the previous "GDDRHammer" and "GeForge" attacks. In most typical server, workstation, and even PC configurations, IOMMU restricts the GPU's access to the CPU's physical addresses, preventing direct memory access. These are the typical DMA-based attacks that the Input-Output Memory Management Unit protects users from. However, the new "GPUBreach" operates differently.

For example, "GPUBreach" exploits memory-safe bugs in the actual GPU driver and corrupts them. When IOMMU confines the GPU's direct memory access to driver-assigned buffers, the new exploit corrupts metadata within these permitted buffers. This causes the driver, which has kernel privileges enabled on the CPU host, to perform out-of-band writes to the buffer, effectively bypassing any protection IOMMU can offer. This logic is built into the kernel by default, as the GPU driver is one of the most trusted components of the operating system. Hence, IOMMU bypass is possible when the metadata is corrupted. Since "GPUBreach" grants an attacker full root privilege escalation, the attack differs significantly from previous rowhammer attacks.

New Rowhammer Attack Puts GDDR6-Powered NVIDIA GPUs at Risk

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 Readies Rubin-based GeForce RTX 60-series with Massive RT Performance Gains

The rumor mill has started grinding about NVIDIA's next-generation gaming GPU, and it looks like NVIDIA does not want to allow the current market environment to get in the way of implementing its product roadmap, with a roughly 2-year GeForce generation product launch cadence. The next-generation GeForce RTX 60-series will be powered by the "Rubin" graphics architecture. "Rubin" already debuted on NVIDIA's bread-winning AI GPU series, and is making its way to GeForce, reports RedGaming Tech.

The first slice of rumors about GeForce RTX 60-series predicts that NVIDIA will stick to a more conservative approach with foundry nodes, and not go with a sub-2 nm nanosheet-based node. GeForce "Rubin" will be built on some variant of the current TSMC 3 nm FinFET node. They need not be the same N3 node that's in use by Apple, Intel, and others; and NVIDIA might collaborate with TSMC on an exclusive variant just the way it created the NVIDIA 4N node, derived from TSMC N5. Chips in the series will follow the numbering scheme "GR20x," with examples being "GR202" for the biggest part powering the flagship product. The 3 nm node will allow NVIDIA to maintain GPU clock speeds ranging between high 2 GHz and low 3 GHz, which is a minor increase over the current "Blackwell."

Lenovo Confirms RTX 5070 Mobile With 12 GB GDDR7 Across Multiple LOQ and Legion Laptops

It recently came to light that NVIDIA may be planning a GeForce RTX 5070 Mobile GPU with an upgrade to 12 GB of GDDR7, 4 GB more than the currently available 8 GB models, with both Lenovo Yoga and ASUS laptops appearing to be planning new models with the new GPU configuration. While there still hasn't been any confirmation from NVIDIA itself, another handful of Lenovo laptop listings have shown up, lending credence to the earlier claims—this time in official, public Lenovo specifications documentation.

According to Lenovo's spec sheet for the Legion Pro 5 16ADR10, the new laptop will pair an AMD Ryzen 9 8945HX with 32 GB of DDR5-5200, a 1 TB PCIe 4 ×4 SSD, and NVIDIA's new GeForce RTX 5070 12 GB. The GPU itself looks largely unchanged from the version in Lenovo's other Legion laptops, with the exception of a much lower boost clock in the Legion Pro 5. Despite having the same 115 W TGP as the LOQ models with the 5070 12 GB, Lenovo has rated the boost clocks of the RTX 5070 12 GB in the Legion Pro at 1425 MHz, compared to 2347 MHz in the LOQ models, which are identical to the original RTX 5070 8 GB. There have been murmurs of an April release for these laptops, but those are unconfirmed so far.

NVIDIA Launches RTX PRO 4500 Blackwell Server Edition GPU

NVIDIA has added another graphics card to its server lineup, this time in the form of a passively cooled, single-slot RTX PRO 4500 Blackwell Server Edition GPU. The company positions this release as a highly efficient GPU for compute-dense environments. It comes with 10,496 CUDA cores, 82 Ray Tracing cores, and 32 GB of GDDR7 memory running on a 256-bit bus, providing 800 GB/s of memory bandwidth, all within a total graphics power of 165 W. This specification is similar to the current RTX PRO 4500 Blackwell with an active dual-slot cooler but saves a few watts of TGP, as the actively-cooled edition has a 200 W TGP. The difference in TGP is attributed to higher-binned "Blackwell" GB203 dies with better frequency tuning, resulting in a similar performance target for this GPU. This server edition SKU also reduces memory bandwidth, running the 32 GB GDDR7 modules at 25 Gbps effective, while the regular blower-style RTX PRO 4500 Blackwell uses full 28 Gbps modules.

This server edition SKU is designed for server configurations that require hyper-dense setups, where a single-slot solution will be cooled by high-RPM server fans. For example, server farms could install a dozen of these GPUs in parallel within a single system, stacking them as long as there are available PCIe slots. High airflow chassis would push air through the passively cooled GPU shroud, cooling the 165 W TGP. Interestingly, this is not even the most efficient GB203 bin with 10,496 CUDA cores, as NVIDIA offers a GeForce RTX 5090 Mobile GPU SKU with only a 95 W TDP. However, that comes at the cost of some clock speeds, which are still unknown for the newest RTX PRO 4500 Blackwell Server Edition GPU.

SK hynix Showcases AI Memory Leadership at NVIDIA GTC 2026

SK hynix Inc. announced today that it is participating in GTC 2026, held from March 16 to 19 in San Jose, California. NVIDIA GTC is the global AI conference where business leaders and developers gather to share the latest breakthroughs and future trends in AI and accelerated computing. SK hynix memory solutions are designed to minimize data bottlenecks and maximize performance for both AI training and inference in NVIDIA AI infrastructure. Through its participation in GTC 2026, the company plans to demonstrate its competitive edge in memory technology—the core infrastructure of the AI era.

Under the theme "Spotlight on AI Memory," SK hynix will feature an exhibition space dedicated to the AI memory technologies and solutions. The booth will consist of three main areas: the NVIDIA Collaboration Zone, the Product Portfolio Zone, and the Event Zone. The exhibition is designed around interactive content to provide visitors with an intuitive understanding of AI memory technology.

NVIDIA GeForce RTX 5050 9 GB Variant Comes with 130 W TDP

NVIDIA is reportedly preparing to update its GeForce RTX 5050 GPU with a new variant featuring 9 GB of GDDR7 memory. A well-known leaker on X, @kopite7kimi, confirmed that the upcoming card will maintain the same 130 W TDP and thermal envelope as the current GeForce RTX 5050, which has 8 GB of GDDR6 memory. The current RTX 5050 uses 8 GB of 20 Gbps GDDR6 memory on a 128-bit bus, providing 320 GB/s of bandwidth. The new model is expected to use three modules of GDDR7 memory, each with 3 GB of capacity, resulting in a total of 9 GB across a 96-bit memory bus. While the narrower bus decreases the interface width, the switch to new 28 Gbps GDDR7 memory would increase total memory bandwidth to 336 GB/s, a roughly 5% improvement, along with a 12.5% boost in VRAM capacity. The new leak also claims that NVIDIA is using GB206 as a GPU base, while the older RTX 5050 8 GB with GDDR6 used GB207 die. However, the core count remains at 2,560 CUDA cores, suggesting that lower-binned GB206 dies found in RTX 5060 and other mid-range SKUs are repurposed for the new RTX 5050.

NVIDIA's switch to GDDR7 memory likely helps the company manage supply chain procurement better, as memory modules are in short supply. Instead of using four GDDR6 modules with 2 GB capacity each, NVIDIA is switching to three modules of GDDR7 with 3 GB capacity each, reducing the number of memory modules needed for this GPU. Interestingly, memory makers like Samsung, Micron, and SK hynix could now produce more of the 3 GB GDDR7 modules, with GDDR6 being in much shorter supply, forcing manufacturers to focus on the newer memory technology while GPU makers have to adapt. This has resulted in a situation where procuring three modules of 3 GB GDDR7 memory is now easier than finding a sufficient supply of GDDR6 modules, which NVIDIA needs to secure for each RTX 5050 GPU. As we approach Computex 2026, this GPU is expected to arrive sometime during that period.

NVIDIA Reportedly Preparing 9 GB GeForce RTX 5050 Variant with GDDR7

NVIDIA is reportedly planning a second RTX 5050 variant with 9 GB of memory instead of the standard 8 GB, according to leaker MEGAsizeGPU who has a decent track record on NVIDIA leaks. Unlike the existing RTX 5050, which uses 8 GB of 20 Gbps GDDR6 memory on a 128-bit bus for 320 GB/s of bandwidth, the new model is said to adopt 3 GB GDDR7 modules. The updated configuration would total 9 GB across a 96-bit memory bus. While the narrower bus reduces interface width, the move to 28 Gbps GDDR7 would raise total memory bandwidth to 336 GB/s, a roughly 5% increase, alongside a 12.5% boost in VRAM capacity. Core specifications are expected to remain unchanged. The card would continue to use the GB207 GPU, based on the Blackwell architecture, with 2560 CUDA cores and a 130 W TDP. No changes to clock speeds or core counts have been mentioned.

The report also claims NVIDIA is preparing adjustments higher up the stack. A revised GeForce RTX 5060 may transition to the larger GB205 die, currently used in the RTX 5070, in a cut-down configuration. That would reportedly reduce the GPU from 6144 CUDA cores and a 192-bit bus to 3840 cores and a 128-bit interface. Such a move would require board partners to adopt new PCBs compatible with GB205 and standard 8-pin power connectors. According to the leak, these changes are tied to supply constraints affecting lower-tier RTX 50-series GPUs, including the RTX 5050 and RTX 5060. The GeForce RTX 5060 12 GB, which appeared in earlier roadmaps, is now reportedly canceled.

ASUS Raises GeForce RTX 50 Series "Blackwell" Pricing in China, Radeon Pricing Unchanged

ASUS has reportedly updated its product pricing in the Chinese market to reflect the memory component shortage across the semiconductor industry. According to Board Channels, ASUS is adjusting the pricing for the GeForce RTX 50 Series "Blackwell" GPUs with GDDR7 memory, while the pricing strategy for AMD Radeon RX 9000/7000 and other series remains unchanged. At the very top, ASUS is increasing the price of its RTX 5090 D v2 SKUs by about 500 yuan, which is approximately $72 at the time of writing. Other SKUs like the GeForce RTX 5080, RTX 5070 Ti, regular RTX 5070, and RTX 5060 Ti with 16 GB of VRAM are experiencing an increase of anywhere from $14.50 to $45. Interestingly, some older and low-end SKUs like the popular GT 1030 and GT 710 series, which provide basic graphics output for many prebuilt PCs, are also seeing a price increase of up to $8. AMD's Radeon series reportedly remains unchanged, which can be explained by these GPUs already having gone through a price increase cycle and GDDR6 memory being in better supply than GDDR7.
You can check out the entire list with proposed price changes below.

MSI GeForce RTX 5090D v2 Lightning Appears with 24 GB VRAM

The MSI GeForce RTX 5090D v2 Lightning has officially appeared on the Chinese market, according to a Bilibili user named "Hardware Patrick Star." This China-exclusive SKU is an adaptation of the new GeForce RTX 5090D v2 with MSI's overclocking enhancements, allowing the card to reach new performance heights, paired with an all-in-one liquid cooler. The new RTX 5090D v2 retains the GB202 family's compute power, with 21,760 CUDA cores, Blackwell architecture, and a 575 W TGP, but reduces the memory to 24 GB of GDDR7 on a 384-bit bus, compared to the previous 32 GB on a 512-bit bus. MSI noted that the GeForce RTX 5090D Lightning GPU is limited to 1,300 units. However, since the China-exclusive SKU is considered a completely different graphics card, it is possible that there are another 1,300 units for the Chinese market with the RTX 5090D v2 Lightning model. The Bilibili user has SKU number 909, indicating that the rest have been distributed to retail channels.

We reviewed MSI's regular GeForce RTX 5090 Lightning graphics card and found that the GPU can reach up to 1,000 W with an OC BIOS loaded. We observed an average GPU clock speed of 3,218 MHz due to the water cooling. Independent testing might show the RTX 5090D v2 Lightning achieving similar performance levels, but since it's a China-focused SKU, we could only see reviews from Chinese hardware reviewers. Earlier claims by EXPreview indicate that hands-on testing suggests mixed results for buyers. In pure gaming, the RTX 5090D and RTX 5090D v2 cards are often nearly identical, with frame rate differences usually within a percent or two, so most current 4K titles will not noticeably suffer from the missing 8 GB. For AI workloads and model inference, the gap is more significant, with single-digit to low-double-digit performance drops where extra memory is important.

NVIDIA Confirms Supply Constraints May Limit Gaming GPU Availability

NVIDIA CFO Colette Kress confirmed that the gaming sector may struggle during the company's latest Q4 earnings call. In a short but very important note, she stated, "Looking ahead, while end demand for our products remains strong and channel inventory levels are healthy, we expect supply constraints to be the headwind to Gaming in Q1 and beyond." This sentence is rather vague but conveys the message that supply constraints will definitely impact the GeForce RTX 50 series lineup in the current quarter and possibly beyond. NVIDIA's current product inventory is in good shape, meaning that both silicon from TSMC and secured GDDR7 memory are sufficient for the time being, but once the inventory levels start to deplete, availability will become a problem.

Team Green has massive capacity secured at TSMC's facilities for manufacturing "Blackwell" GPUs, meaning that no production issues stem from that end. However, memory makers, with whom NVIDIA collaborates, are supply constrained in delivering their GDDR7 memory solutions, leaving NVIDIA with little to work with outside its high-margin server sector. As NVIDIA supplies its AIC partners with both memory and GPU dies, having no memory modules to bundle with the GPUs becomes a supply bottleneck, leaving the company waiting for weeks without a fresh inventory of memory modules. Hence, NVIDIA now expects that demand will continue to be strong among gamers, but the situation may be getting slightly worse as inventory levels start to deplete.

Micron Reveals 24 Gbit GDDR7 Memory with 36 Gbps Bandwidth

Micron chose to unveil its latest developments in graphics memory via a blog post titled "The new performance bottleneck: How more GPU memory unlocks next gen gaming and AI PCs" which explains that its latest GDDR7 memory will allow for graphics cards with as much as 96 GB of VRAM. Micron isn't the first company with 24 Gbit or 3 GB density GDDR7 memory chips, as Samsung sampled partners back in November of last year.

Samsung's 3 GB GDDR7 chips are also offering 36 Gbps bandwidth, so Micron doesn't stand out here either and this might be why the company snuck out the details of its new product in a blog post rather than putting out a press release. Micron claims that its new GDDR7 chips offer improved efficiency, but doesn't go into any details as to what this means. The rest of the blog post is mostly about the benefits of GDDR7 at higher density and speeds and the benefits it will bring to graphics cards, AI PCs and what not. The chips aren't listed on Micron's website as yet, despite the blog linking to Micron's GDDR7 product page.

Sony Could Push PlayStation 6 to 2028/2029 Due to Rising Memory Costs

Sony is reportedly postponing the launch of its next-generation PlayStation 6 console to 2028 or 2029 due to the massive rise in memory costs. This is reportedly impacting decision-making at the top of the company unit responsible for PlayStation strategic planning. Rising memory costs and a tight semiconductor supply chain are the core reasons behind any changes. Initially, we expected to see the wave of next-generation consoles in 2027, as AMD recently confirmed Microsoft's next-generation Xbox arriving in 2027. However, Sony could be abandoning the 2027 release window and pushing its PlayStation 6 console to a late 2028 or even 2029 launch, indicating that the company is not capable of securing sufficient DRAM allocation from its partners. This could leave Microsoft's new Xbox console as superior for an entire year or even two before Sony could join the next-generation console wars.
BloombergSony Group Corp. is now considering pushing back the debut of its next PlayStation console to 2028 or even 2029, according to people familiar with the company's thinking. That would be a major upset to a carefully orchestrated strategy to sustain user engagement between hardware generations.

NVIDIA's Chinese Market Exclusive RTX 6000D 84 GB Model Torn Down by Vlogger

NVIDIA continues to gingerly navigate trade restrictions that affect shipments of high-end graphics cards and AI accelerators into China. Over the past couple of years, region-specific "nerfed" GPU specifications and reduced allocations of VRAM have produced a good number of "lesser" gaming and professional components. For example, a November 2025 leak suggested the readying of an appropriately "reduced" RTX 6000D Blackwell model—seemingly based on Team Green's "unpasteurized" Western market variant. Despite recently arriving without a "PRO" designation, Chinese AI companies and well-heeled PC enthusiasts will be receiving a highly capable bit of hardware. The passively-cooled NVIDIA RTX 6000D card carries 84 GB of GDDR7 video memory; representing a slight downgrade over its non-nerfed sibling's provision of 96 GB (on a 512-bit bus). A video teardown—courtesy of the GINNSOD Bilibili channel—shows an incredibly dense module layout on the card's PCB, consisting of 28 individual 3 GB chips.

This clever clamshell layout surrounds a Team Green "GB202-891-KA-A1" GPU die. Aptly, GINNSOD's presenter refers to the TechPowerUp GPU database for key points of interest—notably, the card's 448-bit memory bus, 19968 shader unit count, and 2430 MHz boost clock. The reference board design is listed as having a 600 W TDP rating, but yesterday's presentation shows only a maximum of 419 watts being sipped by the lab's test subject. International observers have noted that the gory teardown shows a PCB design reminiscent of the RTX PRO 6000 Blackwell model's internal setup, though the latter enjoys having all of its VRAM module bays filled up (with thirty-two 3 GB chips). By contrast, the RTX 6000D's board features four unoccupied spaces. As a system integrator, GINNSOD offers aftermarket liquid-cooled modifications/upgrades—the latter half of the teardown vid delves into the fitting of a custom-made solution. VideoCardz noted that the company has listed a ~$26,000 workstation system armed with twin RTX 6000D cards and an AMD Threadripper Pro CPU. Said pre-built computer can be seen at various points in GINNSOD's showcase.

Leak Suggests Generous Memory Specs for PlayStation 6 Console & Handheld Sibling

Following a mid-January PlayStation 6 chipset leak, Kepler L2 has predicted memory configurations for Sony's next-gen home console (sometimes codenamed "Orion"), as well as a futuristic handheld system. This past weekend, the experienced tech soothsayer outlined key points of interest via the NeoGAF discussion board. In a thread mostly focused on a "next-gen Xbox," a slight deviation into PlayStation territories produced a notable prediction of "up to 30 GB of GDDR7 memory" for the PS5's rumored successor. When answering queries about this impressive dollop of system RAM, Kepler L2 believes that this maximum provision is achievable through the use of 3 GB GDDR7 memory modules in an unusual clamshell configuration.

Commenters did not think that this would be possible with "only 4 GB (GDDR7) modules on a 160-bit bus." In theory, a total bandwidth of 640 GB/s could be enabled, even with a reduced bus (from 256-bit down to 160-bit)—leading to an 11% higher figure than the PS5 Pro's main transfer credential (576 GB/s). Additionally, Kepler L2 believes that the Sony PlayStation engineering team will be bunging (up to) 24 GB of memory into the speculated "PS6 Portable" (aka "Project Canis") handheld gaming console. Previous leaks have suggested the implementation of premium low-power DRAM (LPDDR5X), an expected selection for manufacturers of high-end mobile devices. This memory spec could place the PS6 handheld within the region of ASUS ROG's Xbox Ally X model. Currently, the main caveat of generous/impressively large pools of system memory boils down to pricing and availability. An ongoing supply crisis has already driven fresh rumors of Sony delaying its next-gen gaming systems beyond a proposed 2027 launch window.

Next-Generation Xbox is Windows 11 PC/Console Hybrid for Gaming and Productivity

Microsoft's next-generation Xbox console is reportedly taking an unconventional route by running on a customized version Windows 11 OS instead of the specialized console OS that typically powers these devices. According to Windows Central, the system will function essentially as a gaming PC that boots into an Xbox interface by default. This UI is likely similar to what the current Xbox Full Screen Experience looks and feels like, and will likely provide the same performance boost. We have already seen that Xbox FSE mode brings about a 9.3% reduction in RAM usage and about an 8.6% higher FPS due to the smaller system overhead. Users could exit that interface to access the full Windows 11 operating system, meaning the hardware would support Steam, EPIC, and other competing game stores, as well as standard PC applications alongside Xbox games.

This is Microsoft's first radical departure from the walled-garden approach that has defined console gaming for decades. What it could translate to is the first hybrid system that serves multiple purposes, from traditional gaming to running productivity suites of Microsoft 365 apps like Word, Excel, and others, all from the same system. Teams from the Windows and Xbox divisions are reportedly collaborating closely to adapt the operating system for living room use. Microsoft is also working with hardware partners like ASUS to create multiple devices at different price points rather than releasing a single standard console. Plans for a first-party handheld device are still under consideration, though the traditional console appears to be the main focus.

No NVIDIA GeForce RTX 50 "SUPER" GPUs This Year, RTX 60-Series Also Pushed Back

Artificial Intelligence may be eating the world of software now, but gamers are suffering. According to The Information, NVIDIA has reportedly entirely postponed the launch of its GeForce RTX 50 "SUPER" refresh, as the company's executives are prioritizing AI accelerators over the gaming sector, which consumes precious cutting edge GDDR7 memory. The GeForce RTX 50 "SUPER" refresh was originally scheduled for an announcement at CES 2026, with shipping in Q1 or Q2 of 2026. However, the GDDR7 memory used in the SUPER lineup was a high-capacity 3 GB version, which NVIDIA managers in December deemed too important for gamers, postponing the refresh entirely.

The "SUPER" series was planned with denser GDDR7 memory modules, offering 3 GB of capacity per chip, increasing the memory configuration of the standard GeForce RTX 5070, RTX 5070 Ti, and RTX 5080. Initially, the RTX 5070 SUPER was planned with an upgrade to offer 18 GB, while the RTX 5070 Ti SUPER and RTX 5080 SUPER would each provide 24 GB of GDDR7 memory. As NVIDIA's AI GPU portfolio also uses the high-density GDDR7 memory, like the RTX PRO 6000 "Blackwell" and "Rubin CPX" the company has decided to instead prioritize this high-margin business, leaving gamers with inflated prices of the regular GeForce RTX 50-series.

AMD Confirms Steam Machine in Early 2026, Xbox SoC Powered by RDNA 5 in 2027

AMD posted its record fourth quarter revenue of $10.3 billion in 2025, and during the earnings call, the company issued some guidance on the upcoming product portfolio. During the call, AMD confirmed that Valve's Steam Machine is on track and shipping early this year, while its custom SoC division that designs processors for PlayStation and Xbox consoles will deliver an RDNA 5-based SoC for the next-generation Xbox console. While the Steam Machine specifications are confirmed, Xbox "Magnus" SoC is still largely a collection of rumored specifications. The "Magnus" SoC is rumored to feature the largest APU ever designed for a consumer console, with a 408 mm² chiplet design. Of this, 144 mm² is dedicated to the SoC built on TSMC's N3P node, while the GPU occupies 264 mm². The AMD chip is expected to include up to 11 CPU cores—three Zen 6 and eight Zen 6c—alongside a substantial GPU setup with 68 RDNA 5 compute units, four shader engines, and at least 24 MB of L2 cache. Memory might expand to 48 GB of GDDR7 on a 192-bit bus. A dedicated NPU is rumored to offer significant on-device AI performance, with reports suggesting up to 110 TOPS.
Dr. Lisa Su, AMDFor 2026, we expect semi-custom SoC annual revenue to decline by a significant double-digit percentage as we enter the seventh year of what has been a very strong console cycle. From a product standpoint, Valve is on track to begin shipping its AMD-powered Steam Machine early this year, and development of Microsoft's next-gen Xbox featuring an AMD semi-custom SoC is progressing well to support a launch in 2027.

NVIDIA GeForce RTX 5090/5080 FE SKUs Sell Out in Few Minutes on Company Marketplace

NVIDIA briefly updated its online marketplace with the latest stock of GeForce RTX 5090 and RTX 5080 Founders Edition SKUs, available at MSRP pricing. However, the entire stock was depleted in just minutes after the initial listing, suggesting that gamers have been trying to get their hands on MSRP-priced GeForce RTX 5090 and RTX 5080 GPUs but were able to do so only for a few minutes as the listing disappeared shortly thereafter. The flagship GeForce RTX 5090 Founders Edition was listed at its Euro-designated MSRP of €2,099, while the high-end GeForce RTX 5080 Founders Edition GPU was listed at its €1,059 MSRP. Both GPUs were listed at 10:01 am, but the RTX 5090 FE was gone by 10:08 am, just seven minutes after listing. For the RTX 5080 FE, the listing disappeared at 10:20 am when the stock was gone. The mid-range RTX 5070 FE was also listed but stayed online for a few hours before disappearing.

For buyers interested in purchasing these GPUs at MSRPs, there is virtually no option besides the NVIDIA Marketplace. Any third-party storefronts, where NVIDIA's official AICs are listing these GPUs, are posting listings that go well above €3,000. The card rarely dips below that price, and the median price listing, according to ComputerBase, suggests that the number now floats at €3,566. Any listing near MSRP is a definite premium for gamers who want to get their hands on an already expensive GPU, meaning that they don't have to pay for inflated pricing caused by the GDDR7 memory shortage and NVIDIA's current production tactics. There haven't been any NVIDIA Marketplace updates since the beginning of December, meaning that nearly two full months have passed since the last "drop."

SK hynix Announces FY25 Financial Results, Posts Record-High Results and Delivers Highest Shareholder Returns

SK hynix Inc. announced today that it has achieved record financial results in 2025—97.1467 trillion won in revenue, 47.2063 trillion won in operating profit (with an operating margin of 49%), and 42.9479 trillion won in net profit (with a net margin of 44%). The results significantly exceeded the previous record set in 2024. The annual revenue increased by more than 30 trillion won while the annual operating profit nearly doubled year-on-year, marking the highest annual performance in the company's history.

Growth momentum accelerated further in the fourth quarter. In addition to HBM, demand on conventional memory solutions for servers increased sharply, to which SK hynix responded proactively. As a result, the company achieved record-high quarterly performance across all three indicators, with revenue rising 34% to 32.8267 trillion won, operating profit surging 68% to 19.1696 trillion won, and operating margin reaching 58% quarter-on-quarter.

ZOTAC Korea Notifies Local Customers About "Staggering" GeForce RTX 50-series Price Hikes

Lately, North American customers have shared abject feedback on the internet, regarding frustrating Zotac GeForce RTX 50-series graphics card purchases. Last week, the reported end of NVIDIA's Open Price Program (OPP) likely caused widespread panic amongst hardcore and enthusiast PC DIY communities, leading to swift runs—in-person and online—on multiple retail outlets. The ongoing computer memory crisis has driven up prices across many market sectors—in particular, sharp inclines have affected supplies of GDDR6 and (even more costly) GDDR7 VRAM modules. Sensing a grim near future outlook, Zotac's South Korean office has briefed its local customer about an incoming storm of preposterously epic proportions.

Western news sites were alerted to this sympathetic announcement by harukaze5719—the tenured tracker of tech industry developments summarized part of Zotac SK's message: "(this) current situation threatens the very existence of (add-in-board partners) AIBs and distributors." Earlier today, the South Asian manufacturer's local office issued a major warning, via its official Tagtag Mall account. Starting off, Zotac Korea announced the end of a very minor percentage discount promo: "we recently announced a 2% savings rate. I'm so sorry to be giving you another notice of your savings after only one month. The current situation is so serious that it is worrisome for the future existence of graphics card manufacturers and distributors. They announced that memory supply will not be sufficient and that GPU supply will also be reduced. Additionally, some models will reportedly not be available for some time. Personally, I'm also concerned that it may be impossible for Samsung to secure a smooth supply of GPUs beyond those it can manufacture."

Colorful Expands iGame RTX 5070 Mini OC Series with White Edition

Colorful has expanded its compact iGame Mini graphics cards lineup with a new white variant, the iGame GeForce RTX 5070 Mini W OC 12 GB, following the launch of the black Mini OC models late last year. The new card keeps the same SFF-focused design with a compact 180 × 123 × 39.8 mm form factor and true 2-slot thickness. Specifications remain unchanged compared to the original iGame RTX 5070 Mini OC. Alongside 12 GB GDDR7, the card RTX 5070 GPU with 6,144 CUDA cores is clocked at 2325 MHz base and 2557 MHz boost, indicating the same factory overclock as the earlier model.

Display outputs include three DisplayPort 2.1b connectors and one HDMI 2.1b port. Cooling is handled by a single axial fan with auto-stop technology, similar to the black Mini OC version. The Colorful iGame RTX 5070 Mini W OC draws power from a 16-pin 12V-2x6 connector rated for 250 W input. Aside from the new white shroud and backplate, the iGame GeForce RTX 5070 Mini W OC appears to be a cosmetic update, targeting compact and white-themed SFF builds. Colorful has not yet disclosed pricing or availability.

NVIDIA's GPU Memory Bundles Cost Less Than AMD's, Note AIB Partners

Taiwanese media outlet Benchlife has reportedly obtained information about the pricing of AMD and NVIDIA GPU and memory bundles shipped to their respective add-in-board partners. According to their findings, both AMD and NVIDIA have increased the prices of GDDR6 and GDDR7 memory to reflect the current industry-wide DRAM shortage. However, the report highlights that NVIDIA's pricing for GDDR6 and GDDR7 memory is significantly lower than AMD's. As AMD and NVIDIA bundle their GPU dies with GDDR6/GDDR7 memory, they can set the pricing of these bundles based on several factors. These factors include the cost at which they acquire the DRAM and the desired profit margin for the GPU dies. Typically, the memory is paired with the GPU die, which also influences the final pricing strategy.
BenchlifeOn January 16th, NVIDIA officially notified its AIC partners that all bundled GDDR6 and GDDR7 memory packages are now increasing in price. We have been informed of the price increase through familiar AIC partners, but we won't specify the exact price here. What we can say is that NVIDIA's new prices are still lower than those announced by AMD to its AIB partners.

NVIDIA Reportedly Ends GeForce RTX 5070 Ti Production, RTX 5060 Ti 16 GB Next

NVIDIA has reportedly stopped production of its mid-range GeForce RTX 5070 Ti graphics card, according to ASUS, one of the largest PC manufacturers. ASUS shared exclusive information with Hardware Unboxed, revealing that NVIDIA has created a supply shortage for the GeForce RTX 5070 Ti, effectively ending its production. As one of NVIDIA's largest AIB partners, ASUS has learned that the model is essentially in end-of-life status. The company noted the supply shortage and offered no further comments on availability or potential improvements in supply. Since retailers and manufacturers can't source enough stock of the GeForce RTX 5070 Ti GPUs, the shortage has led to significant price increases for these models and suggests that the future availability of this GPU SKU might be nonexistent.

Update 2, January 16, 08:30 UTC: ASUS issued an official public statement regarding the "incomplete" information about the RTX 5070 Ti and RTX 5060 Ti EOL.
ASUSWe would like to clarify recent reports regarding the ASUS GeForce RTX 5070 Ti and RTX 5060 Ti 16 GB. Certain media may have received incomplete information from an ASUS PR representative regarding these products. The GeForce RTX 5070 Ti and GeForce RTX 5060 Ti 16 GB have not been discontinued or designated as end-of-life (EOL). ASUS has no plans to stop selling these models. Current fluctuations in supply for both products are primarily due to memory supply constraints, which have temporarily affected production output and restocking cycles. As a result, availability may appear limited in certain markets, but this should not be interpreted as a production halt or product retirement. ASUS will continue to support the GeForce RTX 5070 Ti and RTX 5060 Ti 16 GB and is working closely with partners to stabilize supply as conditions improve.

NVIDIA Reportedly Cuts AIC Partner GPU Supply by up to 20%

NVIDIA's shipments to add-in-card (AIC) partners may be 15-20% smaller, according to MEGAsizeGPU on X, who has accurately predicted NVIDIA's lineups in the past. According to that information, NVIDIA's AICs will have to deal with a 15-20% supply cut coming from the mothership. GPU makers like AMD and NVIDIA supply their AICs with raw GPU dies packed with GDDR memory as a kit to put on a customized PCB. AICs are only responsible for engineering and integration, not sourcing these GPU dies, as GPU makers used to provide them in quantities they desired. However, it looks like NVIDIA has decided that the current volume is scheduled for a cut, as up to one-fifth of the AIC GPU shipment will be terminated.

As to why, how, and what of the situation, we only get the what. The reasoning behind the decision and the potential impact on aftermarket pricing is still unknown. Perhaps NVIDIA is reusing its TSMC capacity of consumer "Blackwell" and redirecting it into server-grade "Blackwell" chips like B200 and B300. Additionally, the post also notes that the single reassuring thing is that NVIDIA still bundles GDDR7 memory with its cards that it supplies to AICs, which was at one point a topic of consideration for a breakup. There was a rumor that manufacturers would have to source their own GDDR7 chips from SK Hynix, Samsung, and Micron, in hopes of getting sufficient capacity. However, that has been debunked, and thankfully NVIDIA still provides a GPU and memory kit.

Return to Keyword Browsing
Nov 17th, 2026 20:51 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts