Team Green's CUDA 12.8 release notes have revealed upcoming changes for three older GPU architectures—the document's "Deprecated and Dropped Features" section outlines forthcoming changes. A brief sentence outlines a less active future for affected families: "architecture support for Maxwell, Pascal, and Volta is considered feature-complete and will be frozen in an upcoming release." Further down, NVIDIA states that a small selection of operating systems have been dropped from support lists, including Microsoft Windows 10 21H2 and Debian 11.

Refocusing on matters of hardware—Michael Larabel, Phoronix's editor-in-chief, has kindly provided a bit of history and context. "Four years ago with the NVIDIA 470 series was the legacy branch for GeForce GTX 600 and 700 Kepler series and now as we embark on the NVIDIA 570 driver series, it looks like it could end up being the legacy branch for Maxwell, Pascal, and Volta generations of GPUs." Larabel and other industry watchdogs reckon that the incoming "Blackwell" generation is taking priority, with Team Green likely freeing up resources and concentrating less on taking care of decade+ old hardware. VideoCardz believes that gaming GPU support will continue—at least for Maxwell (e.g. GeForce GTX 900) and Pascal (GeForce GTX 10 series)—based on a playtesting of the toolkit's latest set of integrated drivers (version 571.96).

HXL/9550pro and HardwareLuxx's Andreas Schilling shared evidence of the ASUS PCIe Q-Release Slim—the manufacturer's latest ejection mechanism—causing damage to graphics card PCIe connectors. Recent feedback suggests that repeated usage can scrape or grind off material present on a card's interface. HXL gathered critiques from multiple sources (owners of Intel 800 and AMD 800 series boards), and linked a relevant Bilibili video. The footage presents a damaged GALAX RTX 4070 Ti HOF OC LAB model, following sixty quick release cycles—paired with a ROG Strix B850-A Gaming Wi-Fi S motherboard. Tony Wu, ASUS China's general manager, eventually weighed in on community discussions—stating that he will investigate this matter and report back with his findings.

Schilling expressed similar frustrations—his chosen platform is producing unwanted results: "I'm not happy with the solution either. We use the Strix X870E-E Gaming for testing the graphics cards. So I have had to remove graphics cards from the slot a few dozen times. This didn't always go smoothly and very often the (GeForce RTX 5090) card got stuck in the slot. First damage visible." Press outlets have picked up on the recent surge in Q-Release Slim user feedback—several publications have gathered additional examples of the new mechanism inflicting damage on a variety of graphics card models.

Samsung Electronics Co., Ltd. today announced the Galaxy S25 Ultra, Galaxy S25+, and Galaxy S25, setting a new standard towards a true AI companion with our most natural and context-aware mobile experiences ever created. Introducing multimodal AI agents, the Galaxy S25 series is the first step in Samsung's vision to change the way users interact with their phone—and with their world. A first-of-its-kind customized Snapdragon 8 Elite Mobile Platform for Galaxy chipset delivers greater on-device processing power for Galaxy AI plus superior camera range and control with Galaxy's next-gen ProVisual Engine.

"The greatest innovations are a reflection of their users, which is why we evolved Galaxy AI to help everyone interact with their devices more naturally and effortlessly while trusting that their privacy is secured," said TM Roh, President and Head of Mobile eXperience Business at Samsung Electronics. "Galaxy S25 series opens the door to an AI-integrated OS that fundamentally shifts how we use technology and how we live our lives."

The RTX 50-series gaming GPUs have the gaming community divided. While some appreciate the DLSS 4 and MFG technologies driving impressive improvements in FPS through AI wizardry, others are left disappointed by the seemingly poor improvements in raw performance. For instance, when DLSS and MFG are taken out of the equation, the RTX 5090, RTX 5080, and RTX 5070 are around 33%, 15%, and 20% faster than their predecessors respectively in gaming performance. That said, VideoCardz has tapped into its sources, and revealed the 3DMark scores for the RTX 5090 GPU, and the results certainly do appear to exceed expectations.

In the non-ray traced Steel Nomad test at 4K, the RTX 5090 managed to score around 14,133 points, putting it roughly 53% ahead of its predecessor. In the Port Royal test, which does utilize ray tracing, the RTX 5090 raked in 36,667 points - a 40% improvement over the RTX 4090. The results are much the same in the older Time Spy and Fire Strike tests as well, indicating at roughly a 31% and 38% jump in performance respectively. Moreover, according to the benchmarks, the RTX 5090 appears to be roughly twice as powerful as the RTX 4080 Super. Of course, synthetic benchmarks do not entirely dictate gaming performance, and VideoCardz clearly mentions that gaming performance (without MFG) will witness a substantially more modest improvement. There is no denying that Blackwell's vastly superior memory bandwidth is helping a lot with the synthetic tests, with the 33% extra shaders doing the rest of the work.

During a recent press event in South Korea, NVIDIA addressed concerns about power connector safety for their upcoming RTX 5090 graphics card. The new GPU will consume 575 watts of power, marking a massive 225-watt increase from its predecessor, the RTX 4090. The previous generation RTX 4090 faced significant issues with melting 12VHPWR power connectors, especially with third-party adapters, where incomplete connections led to overheating and connector damage. When questioned about potential risks with the RTX 5090's higher power draw, NVIDIA representatives stated they've implemented an updated 12V-2×6 power connector across the RTX 50 series. Unlike the 12VHPWR 16-pin connector, the new 12V-2x6 has sense pins having recessed further back to ensure proper contact before the GPU can request higher power outputs.

"It is expected that such issues will not occur with the RTX 50 series," a company representative explained during the Q&A session. "After about two years, we believe these problems have been resolved." While the company maintains that user error was the leading cause of failures, the extended timeframe required for developing and shipping revised connectors raised questions about the initial design's reliability. Despite NVIDIA's assurances, the RTX 5090's exceptional power requirements could potentially amplify any unexpected technical issues. The company's previous experience showed that problems became more pronounced in cards with higher power demands, as demonstrated by the RTX 4090 having more incidents than the lower-powered RTX 4080 series. So, more power means more trouble, but the company has worked on it to ensure no future problems arise.

On Tuesday, the newly inaugurated United States president, Donald Trump, announced a massive AI infrastructure expansion in the US called Stargate Project. Stargate is an idea that brings private investments across the US land, with up to 500 billion US dollars committed to the project over the next four years. This is single-handedly one of the most significant infrastructure projects ever planned, and this time it is all about AI and data centers. The initial phase involves deploying 100 billion US Dollars immediately, while the remaining 400 billion will be deployed periodically over the next four years. OpenAI and SoftBank are leading this project, with Softbank's CEO Masayoshi Son being the project's chairman. Major equity partners include SoftBank, OpenAI, Oracle, and MGX. Major technology partners who will supply the know-how, planning, software, and hardware are Arm, Microsoft, NVIDIA, Oracle, and OpenAI.

Leading the entire operation will be up to OpenAI, who is gaining operational lead in the project, while Softbank oversees financial planning. Interestingly, the buildout has already begun. OpenAI is currently exploring a few sites in Abilene, Texas, which includes ten 500,000 sq. ft. data centers with 20 planned for the future. Interestingly, the infrastructure expansion will most likely be present in every US state that can provide ample land and power capacity. OpenAI is looking for partners to help with the massive data centers' power, land, and construction. The most significant impact of this project will be on the power grid, which will require additional buildout and implementation of small nuclear reactors running locally nearby to satisfy the power draw from hundreds of thousands and even millions of GPUs. OpenAI is praising NVIDIA for its almost decade-long partnership, meaning that most GPUs will likely be NVIDIA-sourced.

Sparkle's Arc B580 TITAN Luna OC graphics card model—featuring an all-white design—was discovered via leaked renders around mid-December. By that point in time, the Taiwanese company had already sent Intel Arc B580 "Battlemage" GPU-based products to market—most notably in the form of its Arc B580 TITAN OC SKU, bearing a signature blue design. W1zzard—TechPowerUp's resident evaluator of GPUs—honored this particular model with two awards: "Highly Recommended" and "Great Value." We expect the newly announced "Luna" (almost all-white) counterpart to perform nigh identically—Sparkle has simply rolled out a different color option for customers who favor pale-tinted PC builds.

The inclusion of a differently shaded I/O ruins the overall effect, but snow-white enthusiasts will be happy to discover that Sparkle has produced a white PCB design for its TITAN Luna lineup. Similarly, certain Chinese brands have consistently delivered many white-hued boards. Western PC gaming hardware enthusiasts have often admired the (more recognized) PowerColor Hellhound "Spectral White" aesthetic. A limited edition GPU holder will be bundled in the Arc B580 TITAN Luna OC's retail package—unfortunately, Sparkle has seemingly settled on including this all-blue accessory.

Another day, another Blackwell performance leak. As we inch closer to their official release, more and more benchmark results keep popping up on the internet. Now, a fresh Geekbench OpenCL listing seemingly sheds light on the performance brought to the table by the RTX 5090 Laptop GPU, housed in a Razer Blade 16 alongside a Strix Point APU. However, be warned - the following results depict the RTX 5090 laptop running at only 1515 MHz, which makes it significantly slower than what it is capable of when allowed to run at full tilt, likely to be somewhere around 2100 MHz.

As such, the RTX 5090 Laptop managed a score of only around 91,000 points - almost half that of its predecessor's typical score of 179,000 points according to Geekbench. This likely indicates that the laptop was running on battery power, which would make sense considering the abysmal result. At least, thanks to this listing, we do get to confirm the specifications of the RTX 5090 Laptop GPU - 82 SMs, which means a total of 10,496 CUDA cores, and 24 GB of GDDR7 VRAM. The RTX 4090 Laptop, on the other hand, packs 9728 CUDA cores, putting the RTX 5090 Laptop ahead by around 7.87%. The 150-watt RTX 5090 Laptop also clocks at 1455 MHz at an 80-watt TGP, and 2040 MHz at 150 watts. This does align with what we are witnessing from the aforementioned RTX 5090 Laptop listing, confirming that it is indeed in some kind of low-power mode.

Recent MAXSUN activities have been very Intel Arc "Battlemage" B-series GPU-focused—most notably, a leaked design that leverages dual M.2 slots. Earlier today, ITHome reported on the fresh retail release of two sort-of-new MAXSUN graphics cards in China (on JD.com)—MS-GT730 PH 4 GB (¥359/~US$49) and MS-GT730 PH 2 GB (¥329/~US$45). These super low budget offerings signal a dip back into technologies of old: NVIDIA's GK208 "Kepler 2.0" GPU—of 2014 vintage. This is a curious choice, given that Team Green's "Kepler" GPU generation has not received new drivers since the release of R470 (back in 2021), and the last security update arrived with NVIDIA's 473.47 WHQL driver.

ITHome reckons that MAXSUN has simply recycled its existing MS-GT730 Heavy Hammer 4G and MS-GT730 Heavy Hammer 2G card designs for an early 2025 market launch—TechPowerUp's GPU database produces another very similar looking model: the GT 730 Transformers III 2G. The online publication reached out to MAXSUN's JD storefront customer service for comment on the apparent refresh of old products. ITHome discovered that the renamed reissues are a result of demands generated by "internal supply chain management." MAXSUN stated that its production batch have changed, and it: "emphasizes that the performance (of their new models) is consistent with the old version(s)."

NVIDIA's deep learning super sampling (DLSS) has undergone many iterations to the current version 4 with the transformer model, delivering new technologies such as DLSS Multi Frame Generation, predicting multiple frames in advance to generate the upcoming frame, and increasing the frame output per second. However, not every NVIDIA GPU generation supports these more modern DLSS technologies. In an interview with Digital Foundry, Bryan Catanzaro, VP of Applied Deep Learning Research at NVIDIA, commented on trickling down some DLSS technologies to older GPU generations. For example, DLSS Ray Reconstruction, Super Resolution, and Deep Learning Anti-Aliasing (DLAA) work on NVIDIA GeForce RTX 20/30/40/50 series GPUs. However, the RTX 40 series carries an exclusive DLSS Frame Generation, and the newest RTX 50 series carries the DLSS Multi Frame Generation as an exclusive feature.

However, there is hope for older hardware. "I think this is primarily a question of optimization and also engineering and then the ultimate user experience. We're launching this Frame Generation, the best Multi Frame Generation technology, with the 50 Series, and we'll see what we're able to squeeze out of older hardware in the future." So, frame generation will most likely arrive on the older RTX 30 series, with even a slight possibility of the RTX 20 series getting the DLSS frame generation. Due to compute budget constraints, the multi-frame generation will most likely stay an RTX 50 series exclusive as it has more raw computing power to handle this technology.

AMD's upcoming Radeon RX 9070 series graphics cards have hit an unexpected roadblock, according to recent reports from PC Games Hardware. Despite physical units already reaching select retailers, the launch appears to be delayed due to ongoing pricing negotiations. Industry insider and forum moderator "pokerclock," known for accurate predictions about NVIDIA's GeForce RTX 50 series, reveals that AMD's initial pricing strategy has created tension with retail partners. While boxes bearing the RX 9070 branding have been spotted in retail channels, disagreements over costs have prevented an official release. The core issue stems from AMD's aggressive pricing approach for both the RX 9070 and RX 9070 XT models. Retailers have pushed back against what they consider excessive wholesale costs, forcing AMD to reconsider its strategy.

The company now faces the complex task of potentially reducing prices while compensating retailers who have already purchased inventory at higher rates. Sources suggest AMD may offer marketing funds or cashback incentives to bridge the price gap, though negotiations have reportedly stalled. For example, we recently reported on the AMD Radeon RX 9070 XT AIB model costing around $549. However, NVIDIA has announced its GeForce RTX 5070 at the same $549 price point, with potentially equal or higher raster, ray tracing, and AI capabilities across the board. For AMD to make the value case, the company would need to undercut NVIDIA's pricing. Until that is resolved, retailers aren't allowed to place RDNA 4 GPUs in general sale yet.

A moderator on the PC Games Hardware (PCGH.de) discussion board had disclosed worrying details regarding stock limitations—presumably affecting the upcoming GeForce RTX 50 series launch in Germany. In turn, this disclosure was picked up by PCGH's new department. The predicted circumstances will—reportedly—make matters most difficult for customers looking to acquire higher-end "Blackwell" GPUs. The forum moderator gathered damning evidence from his network of contacts: "I was able to learn from well-informed dealer circles, the available contingent of graphics cards will be extremely limited! This applies in particular to the GeForce RTX 5090. Accordingly, NVIDIA determines where and who exactly will offer graphics cards at market launch. B2B dealers and the entire local wholesale trade, which primarily also works with business customers, will most likely come away empty-handed."

A bit of humor was sprinkled in with this informative post—the moderator joked about customers resorting to "cheerful" repetitive pressings of their F5 keys. They posit that the online buying experience for flagship Blackwell GPUs will be tiring and frustrating: "...so anyone who wants to get a GeForce RTX 5090 or GeForce RTX 5080 at market launch will have to queue digitally at the end customer dealers together with waiting (private) customers. Scalpers and bots will probably also get involved here. The quantities that can be purchased are likely to be limited to a maximum of one unit." Several stores are listed as being prime sources of stock (see below)—they reckon that the likes of Amazon will be not be receiving initial batches. "Second, third, or even fourth" waves of stock are anticipated, with some retailers set to act as resellers—inevitably opening the door to predicted price gouging. It is not clear whether these alleged restrictions will come into effect in markets beyond German borders—additionally, the VideoCardz insider network has not discovered any behind-the-scenes information regarding Team Green's launch period supply strategy.

PowerColor started its online marketing campaign for new Reaper graphics card family earlier this week—a rendered scythe graphic was posted on social media along with this cryptic message: "The Reaper has arrived. Everything is under your control. Will you be the Reaper or the one reaped?" The Taiwanese graphics cards company has already unveiled its opening salvo of new RDNA 4-based card designs—on the internet and in real life. For example, PowerColor's Radeon RX 9070 XT Reaper model was on display at CES 2025—where TechPowerUp spent a couple of minutes with an SFF-form-factor-friendly demonstration sample. Since then, more photo evidence has been posted on the AMD subreddit—a UK retailer appears to have units in-stock at their warehouse.

Team Red is seemingly operating in silent mode—they have not revealed concrete details about the upcoming launch of Radeon RX 9070 XT and Radeon RX 9070 (non-XT) GPUs. Preliminary specification leaks and photos of boxed retail units have turned up this week—with yesterday's Reddit post indicating that Scan UK has received a big cardboard box containing PowerColor Radeon RX 9070 Reaper cards. Industry watchdogs reckon that AMD is still forming a release strategy—with board partners and retail/e-tail outlets waiting on and seemingly ready to receive new or finalized instructions.

AMD's Radeon RX 6750 GRE GPU is an (RDNA 2) Chinese market exclusive—Board Channels members reckon that this model is facing imminent discontinuation. The "GRE" affix denotes "Golden Rabbit Edition." Western graphics card enthusiasts will be familiar with this slightly odd moniker due to Radeon RX 7900 GRE-based (RDNA 3) cards heading to European and North American territories around early 2024. Sadly, this mid-to-high tier GPU is rumored to be reaching an end-of-life (EOL) phase—the latest speculation now points to Team Red's Radeon RX 6750 GRE being quietly withdrawn from China's PC hardware market.

Two Radeon RX 6750 GRE Navi 22-based SKUs—with 10 GB or 12 GB video memory configurations—were created specifically to fill a gap not occupied by AMD's Radeon RX 7600 XT model. The affordable-tier Golden Rabbit Editions are said to be highly popular with Chinese system integrators and budget-minded gamers, but Team Red and its board partners have (apparently) declared that the 10 GB variant is now discontinued. A Board Channel member posited that an "inventory digestion phase" is already underway, with an end date set in March. AIBs are reportedly waiting on "replacement plans from February to March." The Radeon RX 6750 GRE 12 GB variant is reported to be sticking around for the moment, but a new candidate could be appearing in the near future. VideoCardz believes that AMD is considering a rollout of the—not yet announced and oft-delayed—Radeon RX 7650 GRE as a potential modernized substitute.

At the RTX "Blackwell" Editor's Day during CES 2025, NVIDIA pulled back the curtain on one of its most powerful tools: a dedicated supercomputer that has been continuously improving DLSS (Deep Learning Super Sampling) for the past six years. Brian Catanzaro, NVIDIA's VP of applied deep learning research, disclosed that thousands of the company's latest GPUs have been working round-the-clock, analyzing and perfecting the technology that has revolutionized gaming graphics. "We have a big supercomputer at NVIDIA that is running 24/7, 365 days a year improving DLSS," Catanzaro explained during his presentation on DLSS 4. The supercomputer's primary task involves analyzing failures in DLSS performance, such as ghosting, flickering, or blurriness across hundreds of games. When issues are identified, the system augments its training data sets with new examples of optimal graphics and challenging scenarios that DLSS needs to address.

DLSS 4 is the first move from convolutional neural networks to a transformer model that runs locally on client PCs. The continuous learning process has been crucial in refining the technology, with the dedicated supercomputer serving as the backbone of this evolution. The scale of resources allocated to DLSS development is massive, as the entire pipeline for a self-improving DLSS model must consist of not only thousands but tens of thousands of GPUs. Of course, a company making 100,000 GPU data centers (xAI's Colossus) must save some for itself and is proactively using it to improve its software stack. NVIDIA's CEO Jensen Huang famously said that DLSS can predict the future. Of course, these statements are to be tested when the Blackwell series launches. However, the approach of using massive data centers to improve DLSS is quite interesting, and with each new GPU generation NVIDIA release, the process is getting significantly sped up.

To be honest, it is rather tiresome to look for mini PCs with adequately powerful graphics. Of course, AMD's Strix Halo is all set to change that narrative with the help of its powerful integrated graphics that rivals discrete cards, but those stuck with last-gen systems have no option but to opt for an eGPU. Thankfully, the options are plenty on the market, and prominent mini PC brand GMK has now arrived with its own solution as well.

Dubbed the GMK AD-GP1, the eGPU sports the decently powerful AMD Radeon RX 7600M XT graphics card with a measly 8 GB of GDDR6 memory. While that would have been enough 4 or 5 years ago, 8 GB of VRAM is hardly sufficient for modern titles, especially at higher resolutions. Making matters worse, the GMK AD-GP1 does not feature a replaceable GPU, which means that its buyers will be stuck with the RX 7600M XT that its ships with. That said, the system is undeniably rather compact at 16.3 x 11.0 x 3.9 cm, and sports a visually appealing design.

PowerColor has updated its website with a Radeon RX 9070 XT Red Devil product page—this is the first example of an RX 9000 series model being officially listed alongside their existing selection of (exclusively) AMD GPU-based graphics cards. The Taiwanese brand has not published any technical specifications—Team Red RDNA 4 NDAs are likely still in effect—but a pleasing number of Radeon RX 9070 XT Red Devil promotional images have been uploaded. A limited edition package (with alleged bundled extras) seems to be in the pipeline—VideoCardz has provided visual evidence of a fancy container (see below).

Teaser images appeared online at the start of this year—close-ups of glowing signature red parts were accompanied by an ominous message: "every edge shines like a gem. Every second burns like fire. If power was in your hands, how would you use it?" Days later, TechPowerUp inspected a fully unveiled Red Devil demonstration sample at CES 2025—new Hellhound and Reaper designs were also within reach. PowerColor's freshly uploaded images reveal one major difference—VideoCardz adeptly points out the presence of two 8-pin power connectors on the promos, while the CES example possessed three physical inputs. They theorize that renders of PowerColor's Radeon RX 9070 (non-XT) Red Devil model have appeared on the XT's product page. Beyond discrepancies in connector counts, the overall design matches that of the Las Vegas showcase model.

Supply chain insiders reckon that stock of NVIDIA's "Ada Lovelace" generation GeForce RTX 4070 GPU will be depleted by the end of this month. This revelation emerges on the cusp of Team Green's GeForce RTX 5000 series of "Blackwell" GPUs heading to market (on January 30). The popular Chinese hardware forum—Board Channels—is a reliable source of product "supply and demand" information. Industry insiders believe that Team Green's mid-tier SKU will be completely sold—by the closing of this month—across several unnamed regional markets. NVIDIA's incoming GeForce RTX 5070 SKU (baseline $549 MSRP) appears to be the natural alternative, given that GeForce RTX 4070 SUPER offerings are priced within the range of $550 to $600—according to market research performed by VideoCardz.

A board member disclosed details and prophesized changes up and down NVIDIA's current-gen product stack: "The inventory for NVIDIA RTX 4070 and higher models is expected to be completely cleared by January. Starting from February, major AIC brands are anticipated to have stock remaining only for the RTX 4060 Ti and RTX 4060 series, which still need to be digested. However, since NVIDIA's Q1 begins in February, these two models will essentially enter the final stock phase." VideoCardz anticipates a March launch window for the GeForce RTX 5060 series—this could be well-timed given circumstances surrounding Team Green's (alleged) limiting of GeForce RTX 4060 Ti GPU supplies and the insider's predicted cessation of RTX 4060 (non-Ti) production.

Supermicro, Inc. (SMCI), a Total IT Solution Provider for AI/ML, HPC, Cloud, Storage, and 5G/Edge, is showcasing the latest solutions for the retail industry in collaboration with NVIDIA at the National Retail Federation (NRF) annual show. As generative AI (GenAI) grows in capability and becomes more easily accessible, retailers are leveraging NVIDIA NIM microservices, part of the NVIDIA AI Enterprise software platform, for a broad spectrum of applications.

"Supermicro's innovative server, storage, and edge computing solutions improve retail operations, store security, and operational efficiency," said Charles Liang, president and CEO of Supermicro. "At NRF, Supermicro is excited to introduce retailers to AI's transformative potential and to revolutionize the customer's experience. Our systems here will help resolve day-to-day concerns and elevate the overall buying experience."

ADLINK Technology Inc., a global leader in edge computing, unveiled its new "DLAP Supreme Series", an edge generative AI platform. By integrating Phison's innovative aiDAPTIV+ AI solution, this series overcomes memory limitations in edge generative AI applications, significantly enhancing AI computing capabilities on edge devices. Without increasing high hardware costs, the DLAP Supreme series achieves notable AI performance improvements, helping enterprises reduce the cost barriers of AI deployment and accelerating the adoption of generative AI across various industries, especially in edge computing.

Lower AI Computing Costs and Significantly Improved Performance
As generative AI continues to penetrate various industries, many edge devices encounter performance bottlenecks due to insufficient DRAM capacity when executing large language models, affecting model operations and even causing issues such as inadequate token length. The DLAP Supreme series, leveraging aiDAPTIV+ technology, effectively overcomes these limitations and significantly enhances computing performance. Additionally, it supports edge devices in conducting generative language model training, enabling them with AI model training capabilities and improving their autonomous learning and adaptability.

Yeston's social media accounts have revealed upcoming NVIDIA "Blackwell" GPU-based Sakura Atlantis, Sakura and Game Ace models—a single teaser image shows off three different cards that will eventually contain GeForce RTX 50 series graphics hardware. Yeston has deployed common aesthetic traits across the three ranges—notably white PCBs and I/O brackets. Yeston previously revealed their new Sakura Atlantis shroud design with its recent announcement of an AMD Radeon RX 9070 XT model. Today's product unveiling does not include any information about specific Team Green GPU SKUs, but it would be safe to assume that the upper-echelon models will sport Sakura Atlantis livery.

The older Yeston Sakura and Game Ace designs have been rolled out across a number of current-gen "Ada Lovelace" GeForce RTX 40 series GPUs. A quick analysis of Yeston graphics card product hierarchies shows lower-end GPUs (GeForce RTX 4060 Ti and 4060) sporting the Game Ace shroud. The Sakura design is available across a selection of mid-tier GeForce RTX 4070 Ti and 4070 models. VideoCardz proposes that the Sakura Atlantis cooling solution could be robust enough—based on the Radeon RX 9070 XT's rumored 300 W TBP—to take on GeForce RTX 5080 (360 W) and 5070 Ti (300 W) GPUs. TechPowerUp expects Yeston to make further announcements closer to NVIDIA Blackwell's market debut at the end of this month.

The market for compact gaming handheld is expanding at an unprecedented rate, thanks to the popularity boom that handhelds have received in recent months. The GPD Win Min (2025) is one such offering, sitting alongside the rest of interesting gaming products that GPD is known for. The Win Mini is now available for pre-order on Indiegogo, with a starting price of $769 (backers only), and a retail price of $839. Of course, interested buyers are encouraged to be mindful of the risks associated with crowdfunding campaigns, although GPD has mostly had a complaint-free track record.

Unfortunately, and rather unsurprisingly, the entry-level variant does not ship with the latest AMD Strix Point chips, but rather the older Hawk Point offerings, specifically the Ryzen 7 8840U with the Radeon 780M iGPU with 12 CUs. The higher-tier variants ship with Strix Point APUs, starting at the Ryzen AI 9 HX 365 APU, going all the way up to the 12-core Ryzen AI 9 HX 370 with the powerful Radeon 890M iGPU. To accommodate the new APUs, GPD has reworked the thermals, and has added room for full-sized M.2 2280 SSD drives as well. The product packs hall-effect joysticks, plug-and-play grips, and a bunch of useful ports including USB4, USB 3.2 Type-C, an SD Card slot, an audio jack, as well as a good old USB-A port. The 7-inch 1080p display ramps up to 120 Hz, and the system can be equipped with up to 64 GB of memory and 2 TB of PCIe 4.0 storage, and a 44.24 Wh battery is also present.

NVIDIA's new GB200 "Blackwell" racks are running into trouble (again). Big cloud companies like Microsoft, Amazon, Google, and Meta Platforms are cutting back their orders because of heat problems, Reuters reports, quoting The Information. The first shipments of racks with Blackwell chips are getting too hot and have connection issues between chips, the report says. These tech hiccups have made some customers who ordered $10 billion or more worth of racks think twice about buying.

Some are putting off their orders until NVIDIA has better versions of the racks. Others are looking at buying older NVIDIA AI chips instead. For example, Microsoft planned to set up GB200 racks with no less than 50,000 Blackwell chips at one of its Phoenix sites. However, The Information reports that OpenAI has asked Microsoft to provide NVIDIA's older "Hopper" chips instead pointing to delays linked to the Blackwell racks. NVIDIA's problems with its Blackwell GPUs housed in high-density racks are not something new; in November 2024, Reuters, also referencing The Information, uncovered overheating issues in servers that housed 72 processors. NVIDIA has made several changes to its server rack designs to tackle these problems, however, it seems that the problem was not entirely solved.

Sapphire has half revealed its new PULSE graphics card design—yesterday's social media post provided an early look at the signature black/gray shroud adorned with red lines. This seems to be Sapphire's first push into marketing their upcoming AMD Radeon RX 9070 XT and 9070 (non-XT) custom designs. TechPowerUp handled and inspected a new Sapphire PURE model at last week's CES trade event—this particular sample was included in a round-up of Team Red's partner custom designs—but the newly revealed PULSE card was not showcased.

According to a VideoCardz report, Sapphire's new Radeon RX 9000-series NITRO+ model has been teased in AMD's CES-oriented marketing material. The Hong Kong-based tech graphics card specialist chose to not send their Navi 48-based flagship over to Las Vegas—industry experts believe that AIBs are eagerly waiting on Team Red to settle on and announce an official release date. Marketing campaigns are expected to fully kick in closer to RDNA 4's launch window.

AMD's official announcement of new RDNA 4 generation of GPUs—comprised of Radeon RX 9070 XT and Radeon RX 9070—listed several manufacturing partners. YESTON did not join the likes of ASRock, PowerColor, Sapphire and XFX in showcasing custom designs "in-person" at CES, but their official social media accounts have unveiled a new Team Red GPU-based Sakura Atlantis edition card. YESTON has once again deployed its signature baby blue and pink color scheme—quite refreshing when compared to the slew of black/gray custom designs presented at last week's trade show.

Yeston was the first AMD AIB to publicly reveal Radeon RX 9070 XT's video memory allocation online—their social media post confirms the presence of 16 GB VRAM. Team Red partners in attendance at CES 2025 were not allowed to divulge this information, but a number of press outlets read model/name text on certain sticker attachments—corroborating previous leaks that listed 16 GB. Yeston's post outlines a white-lit breathing effects, as well as support for ARGB lighting. The new Sakura Atlantis shroud design is quite unconventional—breaking away from boxy aesthetics. Flowing lines and curved surfaces attract the eye, along with a pleasant pearlescent finish. Shell and starfish motif stickers adorn the card's three cooling fans. VideoCardz reports that Yeston has provided additional details—their Radeon RX 9070 XT-16G Sakura Atlantis model will feature an all-white PCB design and a white-colored I/O bracket. No surprises here, given the company's past choices.