TechPowerUP NEWS

A new adventure is here, adventurers! Today—on February 10—start your journey in the [Only registered and activated users can see links. Click Here To Register...] Demo, team up with friends, explore brand new levels and mechanics! Today, Owlcat Games, alongside META Publishing and [Only registered and activated users can see links. Click Here To Register...], is proud to announce that the first-ever demo playtest for their puzzle action-adventure title, Biped 2, is now live until February 17. Fans of the original Biped and curious newcomers can access the playtest on the game's [Only registered and activated users can see links. Click Here To Register...]. The adorable bipedalled robots you know and love are back! Biped 2 is the sequel to the exciting [Only registered and activated users can see links. Click Here To Register...] about two charming bipedal robots, Aku and Sila, who rely on each other and embark on incredible adventures side by side. Dive into the groovy co-op action-adventure game full of puzzles, friendship, support, and fun!

Biped 2 raises the stakes with expanded multiplayer, including four-player co-op for the first time, new robots, missions, and stunning new biomes. Players will encounter even more challenging puzzles as they navigate diverse landscapes, either solo or with friends in two-or four-player co-op. After brave bipeds had reactivated all the beacons on Earth, they received a strange signal from a faraway planet. This can mean only one thing—someone is calling out for help! Now, our mechanical friends are determined to prove themselves and help those in need across the Galaxy. Become a hero and save the planet in distress. After all, even the tiniest robots are capable of great deeds![Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

MSI has, quite appropriately, introduced new SHADOW 3X graphics card models in a low-key manner. These stealth-black [Only registered and activated users can see links. Click Here To Register...] and [Only registered and activated users can see links. Click Here To Register...] custom designs have just appeared online via [Only registered and activated users can see links. Click Here To Register...]. We expect to see press material published in the near future, perhaps closer to the launch of NVIDIA's [Only registered and activated users can see links. Click Here To Register...]. MSI's newly discovered product pages place emphasis on the SHADOW 3X's essential features: "a performance-focused design that delivers the gaming experience players want, making it the ideal choice when upgrading or building a gaming rig." [Only registered and activated users can see links. Click Here To Register...] believes that the incoming SHADOW 3X cards will be the first to reach Western markets—previous-gen models have released as Chinese regional exclusives. The [Only registered and activated users can see links. Click Here To Register...] (OC) and [Only registered and activated users can see links. Click Here To Register...] SKUs appear to re-use MSI's familiar [Only registered and activated users can see links. Click Here To Register...], just tinted many shades darker.

The new SHADOW 3X cards seem to utilize a fresh design, with no recycling from a similarly appointed sibling product family. MSI's latest barebones/essential tier brings a minimal set of perks: a sizable heatsink, a triple-fan cooling solution, a single 16-pin power connector and a straightforward shroud and backplate aesthetic. SHADOW 3X eschews anything RGB lighting-related. Potential buyers will not be dealing with fancy vapor chambers or liquid cooling systems. MSI will be offering overclocked editions, as well as standard models that conform to NVIDIA reference specifications. TechPowerUp's well-maintained GPU database will be updated with these new additions.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

IBM and Lenovo today announced at [Only registered and activated users can see links. Click Here To Register...] a planned expansion of their strategic technology partnership designed to help scale the impact of generative AI for clients in the Kingdom of Saudi Arabia. IDC expects annual worldwide spending on AI-centric systems to surpass $300 billion by 2026, with many leading organizations in Saudi Arabia exploring and investing in generative AI use cases as they prepare for the emergence of an "AI everywhere" world.

Building upon their 20-year partnership, IBM and Lenovo will collaborate to deliver AI solutions comprised of technology from the IBM watsonx portfolio of AI products, including the Saudi Data and Artificial Intelligence Authority (SDAIA) open-source [Only registered and activated users can see links. Click Here To Register...] (ALLaM), and Lenovo infrastructure. These solutions are expected to help government and business clients in the Kingdom to accelerate their use of AI to improve public services and make data-driven decisions in areas such as fraud detection, public safety, customer service, code modernization, and IT operations.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Sapphire Technologies has started teasing new PURE series graphics cards; earlier today a [Only registered and activated users can see links. Click Here To Register...] stated that new offerings are: "coming your way." An accompanying image presents two semi-obscured signature white models—press outlets believe that the pair are custom Radeon [Only registered and activated users can see links. Click Here To Register...] and RX [Only registered and activated users can see links. Click Here To Register...] models. Sapphire did not exhibit at CES 2025, but a lonely triple-fan Radeon RX 9070 XT PURE card did make the overseas journey to Las Vegas. [Only registered and activated users can see links. Click Here To Register...], AMD presented a small smattering of RDNA 4 board partner samples at a roundup showcase.

Sapphire started to market its new darker dual-fan [Only registered and activated users can see links. Click Here To Register...] last month, but a series of leaks have ruined the company's (social media-driven) build-up of product anticipation. We have not witnessed any Sapphire Radeon RX 9070 XT or 9070 (non-XT) PURE stock reaching retail storage locations, but lower-end PULSE models have been distributed—[Only registered and activated users can see links. Click Here To Register...], at least East of the Mediterranean. Sapphire's CES demonstration sample sported an almost all-white enclosure, a triple-fan cooling solution and dual 8-pin power connectors. The PURE Radeon RX 9070 XT and RX 9070 (non-XT) models could feature limited RGB lighting—an "ARGB out" connector was present on the back section of Sapphire's presentation card.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AI reasoning models and [Only registered and activated users can see links. Click Here To Register...] are set to transform industries, but delivering their full potential at scale requires massive compute and optimized software. The "reasoning" process involves multiple models, generating many additional tokens, and demands infrastructure with a combination of high-speed communication, memory and compute to ensure real-time, high-quality results. To meet this demand, CoreWeave has launched [Only registered and activated users can see links. Click Here To Register...] instances, becoming the first cloud service provider to make the [Only registered and activated users can see links. Click Here To Register...] generally available. With rack-scale [Only registered and activated users can see links. Click Here To Register...] across 72 NVIDIA Blackwell GPUs and 36 [Only registered and activated users can see links. Click Here To Register...], scaling to up to 110,000 GPUs with [Only registered and activated users can see links. Click Here To Register...] networking, these instances provide the scale and performance needed to build and deploy the next generation of AI reasoning models and agents.

NVIDIA GB200 NVL72 on CoreWeave
NVIDIA GB200 NVL72 is a liquid-cooled, rack-scale solution with a 72-GPU NVLink domain, which enables the six dozen GPUs to act as a single massive GPU. NVIDIA Blackwell features many technological breakthroughs that accelerate inference token generation, boosting performance while reducing service costs. For example, fifth-generation NVLink enables 130 TB/s of GPU bandwidth in one 72-GPU NVLink domain, and the second-generation Transformer Engine enables FP4 for faster AI performance while maintaining high accuracy. CoreWeave's portfolio of managed cloud services is purpose-built for Blackwell. [Only registered and activated users can see links. Click Here To Register...] optimizes workload orchestration by exposing NVLink domain IDs, ensuring efficient scheduling within the same rack. Slurm on Kubernetes (SUNK) supports the topology block plug-in, enabling intelligent workload distribution across GB200 NVL72 racks. In addition, CoreWeave's Observability Platform provides real-time insights into NVLink performance, GPU utilization and temperatures.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

The AMD [Only registered and activated users can see links. Click Here To Register...] and [Only registered and activated users can see links. Click Here To Register...] "Zen 5" processors are due for launch at some point next month, but Team Red's recent-ish [Only registered and activated users can see links. Click Here To Register...] did not include any details regarding prices. Given patterns demonstrated by previous generations of Team Red's popular 3D V-Cache-equipped CPUs, we can safely assume that the incoming duo will demand a premium over the already released [Only registered and activated users can see links. Click Here To Register...] SKU (MSRP: $479). Late last week, momomo_us happened upon speculated price points during a sleuthing session involving a comparison shopping website.

The PCPartPicker's price aggregation engine pulled data from two Newegg listings—now scrubbed from existence—that outlined a cost of $699.99 (plus a $12.41 shipping fee) for the 16-core Ryzen 9 9950X3D model, and $599.99 (plus shipping) for the 12-core Ryzen 9 9900X3D. Fortunately, VideoCardz preserved this information over the weekend. PCPartPicker has removed the aforementioned figures from its price history chart system, and Newegg has delisted the offending pages. The leaked price points align closely with MSRPs set for previous-gen (Zen 4) Ryzen 9 7000-series CPUs: $699 for the [Only registered and activated users can see links. Click Here To Register...], and $599 for [Only registered and activated users can see links. Click Here To Register...]. The de-listed prices could be based on placeholder information—the Ryzen 7 9800X3D launched last November with a generational premium of $30 ([Only registered and activated users can see links. Click Here To Register...]'s original MSRP was $449). AMD has alluded to gaming performance being on a [Only registered and activated users can see links. Click Here To Register...], so the incoming Ryzen 9 9950X3D and 9900X3D models are not expected to surpass the Ryzen 7 9800X3D as "[Only registered and activated users can see links. Click Here To Register...]."[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Valve has updated its Steam platform policies to prevent mobile-style advertising practices from infiltrating the PC gaming market. The new guidelines, recently added to the Steam Terms of Service, explicitly ban any game that requires players to watch or interact with advertisements to progress. Under the revised rules, developers must eliminate any ad systems that force players to engage with promotional content as a prerequisite for gameplay. Games that rely on mandatory ads for rewards or advancement will not be permitted on Steam. While cross-promotional partnerships and product placements remain acceptable, the forced ad model is no longer supported. Mobile games often burden players with unwanted commercial interruptions. Developers are now encouraged to pursue alternative monetization strategies, such as single-purchase models, optional microtransactions, or downloadable content packages.

"If your game's revenue relies on advertising on other platforms, you will need to find a new monetization model in order to release on Steam."—states [Only registered and activated users can see links. Click Here To Register...]. In addition to banning forced advertisements, Valve has introduced a new feature for early-access titles. This functionality displays the time elapsed since the last update, offering players greater transparency regarding game development progress. By drawing a clear line against aggressive in-game advertising, Valve is fighting smartphone-style ads that force players into watching unwanted content. Valve's commitment to ecosystem quality ensures that Steam remains a trusted platform for gamers seeking pure gaming and ad-free experiences. Other platforms are likely to follow suit.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Like the first part, [Only registered and activated users can see links. Click Here To Register...] is a story of people who try to accomplish an impossible mission and survive in the sinister Chornobyl Zone. Chernobylite 2 will be released on [Only registered and activated users can see links. Click Here To Register...], similar to the first part in 2019, which was awarded the Top Early Access Platinum Graduate on Steam. Cole Grey, a former military and security guard, lives in a reality where the economy and industry are based entirely on a substance called Chernobylite. With its help, people have made a real leap in civilization. However, the construction of the utopia has been paid for by massive unemployment, pushing people like Cole to the margins of society. Due to severe financial problems, he joins the mysterious "Planewalker" initiative to provide for his family.

It quickly becomes apparent that there is a shortage of Chernobylite in Cole's universe, and the solution is the morally questionable initiative that involves traveling to parallel worlds in search of the mineral and obtaining it by any means possible. In addition, the Planewalker service requires appropriate genetic modifications. Cole must, therefore, physically absorb the essence of chernobylite and, to increase his physical and intellectual capabilities, lose a part of his humanity. However, seeing no other options for himself, he agrees to the bloody and dangerous service, which results in him ending up trapped in the most sinister place in the world: the Chornobyl Exclusion Zone. There, he meets a Black Stalker, the Zone's dark guide and silent ruler.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

More than 700 games and applications feature RTX technologies, and each week new games integrating NVIDIA DLSS, NVIDIA Reflex, and advanced ray-traced effects are released or announced, delivering the definitive PC experience for GeForce RTX players. Star Wars Outlaws' latest update is out now, adding DLSS 4 with Multi Frame Generation, an upgraded DLSS Frame Generation model for improved performance and reduced VRAM usage, and the new DLSS transformer AI model for DLSS Super Resolution, DLSS Ray Reconstruction, and DLAA, further enhancing your PC experience.

Later today, Avowed launches with DLSS, Reflex, and ray tracing, and by using NVIDIA app's new DLSS 4 overrides, you can further enhance your experience in Avowed by adding DLSS 4 with Multi Frame Generation, and new and improved AI models that upgrade DLSS Super Resolution, DLAA, and DLSS Frame Generation. A new DLSS Frame Generation, Reflex, and ray tracing update for the popular Wuthering Waves launches today, too. And next week, Indiana Jones and the Great Circle introduces DLSS 4 with Multi Frame Generation and DLSS Ray Reconstruction, updates DLSS Super Resolution and DLAA to use our new transformer-based AI model, and adds fully ray-traced shadows from all light sources.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AAEON (stock code: 6579), a leading provider of edge AI platforms, today announced the release of the BOXER-8641AI-Plus, an embedded AI system powered by the NVIDIA Jetson AGX Orin module designed for the smart healthcare space. The BOXER-8641AI-Plus offers a number of advanced features, including an RJ-45 port that supports 10G LAN, two DB-9 ports for RS-232/422/485 (Rx/Tx/RTS/CTS), and a third DB-9 port offering an 8-bit DIO interface. In addition to this, the system contains expansion slots for Wi-Fi/BT and 5G in the form of an M.2 2230 E-Key and an M.2 3052 B-Key, which serve to provide wireless connectivity in addition to its onboard interfaces.

The system is FCC Class B certified, making it compliant with electromagnetic susceptibility (EMS) standards, making it suitable for operation in healthcare settings without electromagnetic interference (EMI) from other equipment. Moreover, the certification means the BOXER-8641AI-Plus meets electromagnetic compatibility (EMC) standards, ensuring it can be deployed without causing electromagnetic interference (EMI) that could affect equipment deployed alongside it.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Back in January, we covered a [Only registered and activated users can see links. Click Here To Register...] about AMD designing its next-generation "Zen 6" CCDs on a 3 nm-class node by TSMC, and developing a new line of server and client I/O dies (cIOD and sIOD). The I/O die is a crucial piece of silicon that contains all the uncore components of the processor, including the memory controllers, the PCIe root complex, and Infinity Fabric interconnects to the CCDs and multi-socket connections. Back then it was reported that these new-generation I/O dies were being designed on the 4 nm silicon fabrication process, which was interpreted as being AMD's favorite 4 nm-class node, the TSMC N4P, on which the company builds everything from its current "Strix Point" mobile processors to the "Zen 5" CCDs. It turns out that AMD has other plans, and is exploring a 4 nm-class node by Samsung.

This node is very likely the Samsung 4LPP, also known as the SF4, which has been in mass-production since 2022. The table below shows how the SF4 compares with TSMC N4P and Intel 4, where it is shown striking a balance between the two. We have also added values for the TSMC N5 node from which the N4P is derived from, and you can see that the SF4 offers comparable transistor density to the N5, and is a significant improvement in transistor density over the TSMC N6, which AMD uses for its current generation of sIOD and cIOD. The new 4 nm node will allow AMD to reduce the TDP of the I/O die, implement a new power management solution, and more importantly, the need for a new I/O die is driven by the need for updated memory controllers that support higher DDR5 speeds and compatibility with new kinds of DIMMs, such as CUDIMMs, RDIMMs with RCDs, etc.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AMD's Radeon RX 9000 series GPUs are expected to come with up to 16 GB of GDDR6 memory. However, AMD is reportedly expanding its RX 9070 lineup with a new 32 GB variant, according to sources on Chiphell. The card, speculatively called the RX 9070 XT 32 GB, is slated for release at the end of Q2 2025. The current GDDR6 memory modules used in GPUs carry a capacity of 2 GB per module only, meaning that a design with 32 GB of VRAM would require as many as 16 memory modules on a single card. No 2 GB+ GDDR6 memory modules are available, meaning that the design would require memory module installation on both the front and back of the PCB. Consumers GPUs are not known for this, but it is a possibility with workstation/prosumer grade GPUs employing this engineering tactic to boost capacity,

While we don't have information on the GPU architecture, discussions point to potential modifications of the existing Navi 48 silicon. This release is positioned as a gaming card rather than a workstation-class Radeon PRO 9000 series product. AMD appears to be targeting gamers interested in running AI workloads, which typically require massive VRAM amounts to run locally. Additionally, investing in a GPU with a big VRAM capacity is essentially "future-proofing" for gamers who plan to keep their cards for longer, as recent games have been spiking VRAM usage by a large margin. The combination of gaming and AI workloads may have made AMD reconsider some of its product offerings, potentially giving us the Radeon RX 9070 XT 32 GB SKU. We have to wait for the Q2 to start, and we can expect more details by then.

Update 20:55 UTC: AMD's [Only registered and activated users can see links. Click Here To Register...] debunked rumors of the 32 GB SKU coming to gamers. So, this will not happen. Instead, we could be looking at prosumer oriented AMD Radeon Pro GPU with 32 GB of memory instead.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...] reports that a federal judge has rejected Western Digital's attempt to delay payment of a $553 million patent infringement penalty, giving the storage giant just seven days to secure a bond or pay up. The ruling, issued February 11, comes as Western Digital faces mounting pressure over intellectual property disputes. To be more specific it's about SPEX Technologies' Patent No. 6,088,802, which protects security features in peripheral devices. In October 2023, a California jury ruled that several WD products, including [Only registered and activated users can see links. Click Here To Register...] and My Book drives infringed on this 1997 patent. Judge James Selna later expanded the initial $316 million verdict to include $237 million in interest charges.

Western Digital's upcoming corporate restructuring complicates the situation. The company plans to [Only registered and activated users can see links. Click Here To Register...] by February 21—one to focus on NAND flash memory under the [Only registered and activated users can see links. Click Here To Register...] brand and another to maintain the traditional hard drive business. This pending division has raised questions about how to enforce the damages, with Judge Selna mentioning his "concerns about potential corporate restructuring" because the judgment targets Western Digital Technologies Inc. SPEX has voiced concerns about the company split pointing out in legal papers their lack of insight into which new entity would pay the penalty or if either would have enough money to cover the owed amount.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]—AMD's corporate vice president and general manager of client channel business—has [Only registered and activated users can see links. Click Here To Register...] as a highly important date for next-gen graphics technology. The much-anticipated ([Only registered and activated users can see links. Click Here To Register...]) Radeon RX 9000 series unveiling event has a confirmed time slot on that day: 8 AM EST. The overseer of Ryzen CPU and Radeon GPUs has warded off curious queries from journalists and members of the PC hardware for several weeks, since the conclusion of CES 2025. A confusing early January presentation did not include a segment dedicated to upcoming RDNA 4 products. Online conjecture pointed to Team Red delaying and [Only registered and activated users can see links. Click Here To Register...] the launch of Radeon RX 9070 XT and RX 9070 models.

AMD enthusiasts will breathe a collective sigh of relief, after reading McAfee's announcement: "the wait is almost over. Join us on February 28 at 8 AM EST for the reveal of the next-gen AMD Radeon RX 9000 Series. Get ready to make it yours when it hits shelves in early March. RSVP by subscribing to the [Only registered and activated users can see links. Click Here To Register...]." Insider sources reckon that retailers will have stock on shelves by the [Only registered and activated users can see links. Click Here To Register...] launch day. Mid-to-late January leaks suggested a fairly comprehensive distribution of board partner custom cards [Only registered and activated users can see links. Click Here To Register...]. Alleged specifications and performance results have leaked out over the past month and a half—will AMD (and AIBs) have any surprises lined up for the February 28 event?[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

The wait is finally over. This is Yang Bing, as the creator of [Only registered and activated users can see links. Click Here To Register...], I couldn't be more excited to share some big news—our action RPG will officially launch on PlayStation 5 and PC on May 30, 2025! What started as a one-person dream (me, a computer, and endless late nights) has evolved into a sprawling adventure full of breathtaking visuals and adrenaline-pumping combat. Let me take you behind the scenes to explore the heart, soul, and sweat poured into this game. Whether you're here for the stylish combat or the gripping story, there's plenty to dive into. Let's get started!

A world that lives and breathes
Creating the world of Lost Soul Aside was like painting with imagination on a sci-fi fantasy canvas. You'll traverse landscapes that feel alive: sunlit plains, mysterious ancient ruins, and alternate dimensions that buzz with energy. This world isn't just eye candy—it's filled with secrets, monsters, and unpredictable events that keep you interested. But let's talk about what really defines the experience—combat.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Sapphire is readying multiple Radeon RX 9070 custom designs for next month's launch; official announcements and leaks have revealed upcoming PURE, PULSE and NITRO+ cards. The latter was [Only registered and activated users can see links. Click Here To Register...] press material, but no demonstration sample turned up at last month's Las Vegas AIB roundup. A lone fuzzy low-resolution screengrab showcased the premium Radeon RX 9070 series card's hefty and blocky triple-slot profile. An hour or two ago, Everest (aka Olrak29) uploaded two alleged promotional images to social media. Thankfully, the leaked shots have arrived with decent pixel definition. Sapphire has initiated staggered marketing campaigns for its lower-end and [Only registered and activated users can see links. Click Here To Register...] RDNA 4 cards, so the latest leak has most likely preempted an upcoming official reveal of (possible) NITRO+ [Only registered and activated users can see links. Click Here To Register...] and [Only registered and activated users can see links. Click Here To Register...] cards.

Sapphire reserves the crème de la crème of feature sets for its premium-tier NITRO+ designs. Previous-gen examples were substantial prospects, and the incoming design is touted to continue that dimensional legacy. Well-heeled Team Red enthusiasts will welcome an updated triple-fan configuration and somewhat subtle internal ARGB lighting zones. The leaked card sports a luxurious "champagne gold" tinted shroud and backplate, with an optional detachable panel. The vented side plates sport a lighter-hue, but the elaborate triangular cut-out aesthetic could be off-putting to certain connoisseurs. Press outlets have honed in on the new NITRO+ model's hidden power connector design, potentially leading to a single input—utilizing a 12 or 16-pin interface.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Hewlett Packard Enterprise announced today that it has shipped its first [Only registered and activated users can see links. Click Here To Register...] family-based solution, the NVIDIA GB200 NVL72. This rack-scale system by HPE is designed to help service providers and large enterprises quickly deploy very large, complex AI clusters with advanced, direct liquid cooling solutions to optimize efficiency and performance. "AI service providers and large enterprise model builders are under tremendous pressure to offer scalability, extreme performance, and fast time-to-deployment," said Trish Damkroger, senior vice president and general manager of HPC & AI Infrastructure Solutions, HPE. "As builders of the world's top three fastest systems with direct liquid cooling, HPE offers customers lower cost per token training and best-in-class performance with industry-leading services expertise."

The [Only registered and activated users can see links. Click Here To Register...] features shared-memory, low-latency architecture with the latest GPU technology designed for extremely large AI models of over a trillion parameters, in one memory space. GB200 NVL72 offers seamless integration of NVIDIA CPUs, GPUs, compute and switch trays, networking, and software, bringing together extreme performance to address heavily parallelizable workloads, like generative AI (GenAI) model training and inferencing, along with NVIDIA software applications. "Engineers, scientists and researchers need cutting-edge liquid cooling technology to keep up with increasing power and compute requirements," said Bob Pette, vice president of enterprise platforms at NVIDIA. "Building on continued collaboration between HPE and NVIDIA, HPE's first shipment of NVIDIA GB200 NVL72 will help service providers and large enterprises efficiently build, deploy and scale large AI clusters."[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

During its first post-Western Digital spinoff [Only registered and activated users can see links. Click Here To Register...], SanDisk showed something it has been working on to tackle the AI sector. High-bandwidth flash (HBF) is a new memory architecture that combines 3D NAND flash storage with bandwidth capabilities comparable to high-bandwidth memory (HBM). The HBF design stacks 16 3D NAND BiCS8 dies using through-silicon vias, with a logic layer enabling parallel access to memory sub-arrays. This configuration achieves 8 to 16 times greater capacity per stack than current HBM implementations. A system using eight HBF stacks can provide 4 TB of VRAM to store large AI models like GPT-4 directly on GPU hardware. The architecture breaks from conventional NAND design by implementing independently accessible memory sub-arrays, moving beyond traditional multi-plane approaches. While HBF surpasses HBM's capacity specifications, it maintains higher latency than DRAM, limiting its application to specific workloads.

SanDisk has not disclosed its solution for NAND's inherent write endurance limitations, though using pSLC NAND makes it possible to balance durability and cost. The bandwidth of HBF is also unknown, as the company hasn't put out details yet. SanDisk Memory Technology Chief Alper Ilkbahar confirmed the technology targets read-intensive AI inference tasks rather than latency-sensitive applications. The company is developing HBF as an open standard, incorporating mechanical and electrical interfaces similar to HBM to simplify integration. Some challenges remain, including NAND's block-level addressing limitations and writing endurance constraints. While these factors make HBF unsuitable for gaming applications, the technology's high capacity and throughput characteristics align with AI model storage and inference requirements. SanDisk has announced plans for three generations of HBF development, indicating a long-term commitment to the technology.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AMD's Radeon RX 9070 GPU series is due for release [Only registered and activated users can see links. Click Here To Register...]; a specific date has not been set, but we will likely find out more through official channels at the [Only registered and activated users can see links. Click Here To Register...]. Team Red and its board partners have chosen to remain silent on the subject of RDNA 4's technical makeup; post-CES 2025, hardware news outlets have relied on a steady trickle of Radeon RX 9070 XT and RX 9070-related leaks. Very basic sleuthing pointed to pools of 16 GB VRAM for both models, while insiders kept on mentioning an unannounced "[Only registered and activated users can see links. Click Here To Register...]" GPU. The latest—courtesy of [Only registered and activated users can see links. Click Here To Register...]—seems to confirm that the Radeon RX 9070 XT will utilize the aforementioned new RDNA 4 Navi chip. Earlier today, a screenshot was uploaded to social media—the leaker shared graphics card information displayed in a [Only registered and activated users can see links. Click Here To Register...] (v2.62) session. Despite patch notes not disclosing compatibility, the latest version of GPU-Z is seemingly able to identify key aspects of the alleged "[Only registered and activated users can see links. Click Here To Register...]" sample.

The card's name is obscured, but HKEPC and several press outlets believe that it is the genuine article. The fundamental details appear to be: 16 GB GDDR6 VRAM (Hynix-made), a 256-bit memory bus, 4096 stream processors, and a boost clock reaching a maximum frequency of 3.1 GHz. Older leaks have indicated that the first wave of RDNA 4 cards will make do with PCI-Express 4.0 x16 interfaces, but the GPU-Z screenshot shows a PCI-Express 5.0 x16 bus interface (detection could be bugged). The driver version was identified as Adrenalin 24.30.01.05. The unnamed card appears to feature a steep factory overclock; industry experts reckon that the sample could be a very high-end AIB model. Past reports suggest that PowerColor's [Only registered and activated users can see links. Click Here To Register...] card is [Only registered and activated users can see links. Click Here To Register...]. HKEPC uploaded another incriminating screenshot; showcasing performance results produced by Capcom's [Only registered and activated users can see links. Click Here To Register...]. The test system—featuring Intel's [Only registered and activated users can see links. Click Here To Register...] CPU and 48 GB of RAM—scored 36102 points and achieved a maximum frame rate of 211.71 FPS at 1080p, with "Very High" profile settings. The leaker confirmed that FSR and Frame Generation were enabled during the benchmark session.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

A recently surfaced [Only registered and activated users can see links. Click Here To Register...] has revealed the performance improvements that the GeForce RTX 5070 Ti is likely to bring to the table, and the numbers sure look promising - that is, coming from the disappointment of the GeForce RTX 5080, which manages roughly 260,000 points in the benchmark, portraying a paltry 8% improvement over its predecessor. The GeForce RTX 5070 Ti, however, managed an impressive 248,000 points, putting it a substantial 20% ahead of the GeForce RTX 4070 Ti. Hilariously enough, the RTX 5080 is merely 4% ahead, making the situation even worse for the somewhat contentious GPU. NVIDIA has claimed similar performance improvements in its marketing material, which does seem quiet plausible.

Of course, an OpenCL benchmark is hardly representative of real-world gaming performance. That being said, there is no denying that raw benchmarks will certainly help buyers temper expectations and make decisions. [Only registered and activated users can see links. Click Here To Register...] have hinted at a roughly 10% improvement over its predecessor in raster performance and up to 15% improvements in ray tracing performance, although the OpenCL listing does indicate the RTX 5070 ti might be capable of a larger generational jump, neck-and-neck with NVIDIA's claims. For those in need of a refresher, the RTX 5070 Ti boasts 8960 CUDA cores paired with 16 GB of GDDR7 memory on a 256-bit bus. Like its siblings, the RTX 5070 is also rumored to face "extremely limited" supply at launch. With its official launch less than a week away, we won't have much waiting to do to find out for ourselves.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]