If your workstation build has no budget limits, there's now a way to get 16 TB of NVMe PCIe 4.0 SSD storage on a single M.2 2280 SSD. However, this small storage drive will set you back nearly $16,000, with Amazon listing a brand new Exascend PE4 16 TB M.2 SSD at $15,935. What makes this drive so expensive is the fact that you are getting 16 TB of NVMe SSD storage in a super-dense configuration that fits into a single M.2 slot on a motherboard, requiring no additional drives to achieve this capacity. The Exascend PE4 drive uses a PCIe 4.0 interface and can reach sequential read speeds of up to 3,270 MB/s, while write speeds can reach up to 2,980 MB/s, just shy of 3 GB/s. The drive uses TLC 3D NAND Flash storage modules from an unknown manufacturer.

The drive is designed to withstand a TBW of 16,640 TB written to it and about two million hours of MTBF, meaning its usable life is guaranteed to last through any known task, running 24/7 in systems where reliability is the number one factor. Interestingly, Exascend only offers a five-year warranty but claims its hardware will significantly outlast that. Rigorous testing and massive capacity justify the high price point, so some users might find the need for such a drive. Additionally, idle power consumption remains under 1.3 W, while this SSD can reach up to 7.2 W of active power when under full load, meaning that efficiency for such a drive is rather good.

AMD is preparing an update to its Ryzen 9000 series desktop processor lineup with the introduction of two new models, the Ryzen 7 9750X, and the Ryzen 5 9650X. Both these chips are non-X3D (lack 3D V-Cache), and implement the regular "Zen 5" CCD with 32 MB on-die L3 caches. The two are being designed with increased clock speeds and TDP, and their launch closely follows Intel's recent product stack refresh with the Core Ultra 7 270K Plus and Core Ultra 5 250K Plus.

The Ryzen 7 9750X is an 8-core/16-thread chip with a base frequency of 4.20 GHz with 5.60 GHz maximum boost frequency, a significant increase over the 3.80 GHz base and 5.50 GHz maximum boost frequency of the Ryzen 7 9700X. The 9750X comes with a 120 W TDP out of the box. In comparison, the 9700X comes with 65 W TDP out of the box, and AMD allowed motherboard vendors to provide a BIOS-based 105 W TDP mode that doesn't break warranty, designed to improve boost frequency residency. The 9750X not only comes with increased clocks, but also increases the TDP further, to 120 W from that BIOS-based 105 W TDP mode.

Microsoft Office 2024 Home & Business is $139.97 (regularly $249.99) for a lifetime license on Mac or PC. Subscription fatigue is real, especially when it comes to software you use every day. Microsoft 365 costs $69.99 to $99.99 annually, which means you'll easily spend $200-$400 within a few years just to keep using Word and Excel. The Microsoft Office 2024 Home & Business lifetime license eliminates the subscription cycle entirely. One payment installs the classic desktop apps directly on your Mac or PC with no recurring fees.

Microsoft Office 2024 Home & Business Lifetime License is $139.97 (regularly $249.99).

ASUS today announced the ExpertCenter PN55 Mini PC, a compact Copilot+ PC powered by the latest AMD Ryzen AI 400 Series processors with class-leading multithreaded performance and advanced XDNA 2 NPU delivering up to 55 AI TOPS. Integrated AMD Radeon 800M graphics provides prosumers and content creators with exceptional, incredibly detailed visuals. Despite its small footprint, ExpertCenter PN55 offers dual LAN and up to six USB ports, giving it the flexibility to take on a variety of tasks including AI-accelerated productivity, collaboration, and content creation. It also offers design features that enable tool-less upgrades.

Enhanced productivity with Copilot+
ASUS ExpertCenter PN55 Mini PC is powered by an up to AMD Ryzen AI 9 HX 470 processor featuring up to 55 TOPS of XDNA2 NPU performance. This robust compute engine enables smooth generative AI tasks, allowing users to generate ideas, create content, and upscale images in seconds. Paired with up to 96 GB of DDR5 memory, ExpertCenter PN55 ensures rapid access to AI-accelerated workloads and supports seamless multitasking for everyday productivity tasks such as web browsing, presentations, and content creation. AI PCs accelerate everyday tasks by automating repetitive work, helping users save time and work more efficiently.

AMD today released a statement in China on the scandal involving Chinese notebook OEM Chuwi mislabeling Ryzen 5000 series mobile processor models as Ryzen 7000 series. You can read all about the scandal in our older article, but to summarize, the company was found selling notebooks with cheap Zen 2 Ryzen 5 5500U processors with the processor name strings modified in the BIOS to falsely show Ryzen 5 7430U, a Zen 3 chip released almost three years later. This attempt at deception also covered product marketing, and Ryzen 7000 series case badges on the notebook chassis.

AMD, in its Chinese-language statement to the Chinese press, as reported by Hong Kong-based HKEPC, came down hard on the malpractice by Chuwi. The company said that this behavior by Chuwi was in no form authorized by AMD; that the company has strict and legally-binding agreements with its OEMs over the handling of the AMD brand, product labels, or product promotion; and condemned the behavior, saying that such acts damage consumer confidence in AMD as a brand. It ended the statement saying that the company reserves the right to pursue legal action against those involved.The machine translated statement by AMD follows:

Thermaltake, a leading PC DIY brand for premium hardware solutions, introduces the TR200 Series, a compact microATX lineup designed to deliver high-end hardware compatibility, advanced cooling support, and modern visual customization within a minimalist form factor. The TR200 Series includes the TR200 and TR200 WS Micro Chassis, both available in Black and Snow, offering identical internal specifications with distinct front panel aesthetics to match different setup styles.

The TR200 WS features vertical real wood accents on the front panel, introducing natural warmth and refined texture while maintaining the same airflow performance and hardware compatibility as the TR200, which features a clean mesh front design. This allows builders to choose between a minimalist, modern aesthetic or a warmer, more decorative style that complements contemporary desk setups and living spaces.

Ahead of Death Stranding 2: On the Beach PC launch scheduled for March 19th, Sony has released a new PC version launch trailer called "No Rain, No Rainbow", as well as confirmed that the game will be getting support for NVIDIA DLSS, AMD FSR, and Intel XeSS upscaling and frame generation technologies. Death Stranding 2: On the Beach will also support ray-traced reflections and ambient occlusion, where ray tracing will be used for surfaces like water and other reflections, while ambient occlusion should bring much more realistic shadows due to ambient lighting effects. Unfortunately, Sony did not include ray tracing in the previously released PC system requirements, so it is left to be seen how the Decima engine handles these effects in terms of CPU and GPU requirements. According to Sony, "these additional PC options are aimed at players with powerful hardware that want to push visual fidelity beyond the "Very High" graphics settings the game already offers."

In addition, the PC version of Death Stranding 2: On the Beach will also get uncapped frame rates, more extensive graphics settings options, super-ultrawide (32:9) and ultrawide (21:9) monitor support, full mouse and keyboard support with key binding, spatial sound support, and more. The PC version of the game will also get a new challenge with the "to the wilder" game mode, as well as some other new game features and content additions. Sony also released a new launch trailer for the PC version of the game, which you can check out below.

With the latest Arc 101.8626 WHQL graphics driver, Intel has extended its precompiled shader delivery service to Intel Arc B-series GPUs and Intel Core Ultra Series 3 and Series 2 SoCs with built-in Intel Arc GPUs. This enhancement aims to significantly reduce game loading times. Intel's service gathers game shaders in the company's private cloud infrastructure, where they are processed and precompiled. When you install the Intel Graphics Software App, the service identifies the games you play and downloads the precompiled shaders for those games, using the Intel app as a distribution service and creating a folder with the precompiled shaders. This approach allows games to load much faster, reduces stuttering on the first launch, and automatically updates shaders whenever they are revised, with Intel's service pulling the new shaders into the shared folder on your PC.

Currently, Intel supports 13 games at launch, which on average offer more than twice the loading speed. Depending on the game, players can expect improvements ranging from a 1.3x increase in The Elder Scrolls IV: Oblivion Remastered to up to 21x faster loading times in God of War Ragnarök. The list of supported games also includes Black Myth: Wukong, Borderlands 4, Call of Duty: Black Ops 6, Call of Duty: Black Ops 7, Cyberpunk 2077, Gotham Knights, Hogwarts Legacy, NBA 2K26, Starfield, S.T.A.L.K.E.R. 2: Heart of Chornobyl, and The Outer Worlds 2.

BIOSTAR, a leading manufacturer of edge computing solutions, industrial motherboards, graphics cards, and storage devices, is proud to introduce the new BITWL-IHT, a Thin Mini-ITX industrial motherboard engineered to deliver stable and scalable performance for modern intelligent infrastructure. Built to smoothly run with the Intel Alder Lake-N, Amston Lake, and Twin Lake series processors, the BITWL-IHT is designed to address the growing demand for reliable edge platforms across AIoT, industrial automation, HMI, kiosk, digital signage, and distributed edge computing environments where long-term stability and flexible integration are essential.

At its core, the BITWL-IHT supports Intel Processor N150, Intel Atom x7213E, and Intel Core i3-N305, offering adaptable processing performance across a wide spectrum of embedded and edge workloads. Combined with support for DDR5-4800 SO-DIMM memory up to 16 GB, the platform ensures responsive multitasking, efficient data handling, and improved power efficiency for continuous industrial operation. Its compact 170 mm x 170 mm slim Mini-ITX form factor enables streamlined system integration in space-constrained environments, making it ideal for slim kiosks, panel PCs, and embedded control systems.

Microsoft has finally released a second DirectX Ray Tracing (DXR) functional specification file that outlines what its ray tracing pipeline is expected to look like, the goals the company is pursuing, and what the technology does behind the scenes. In the original file, Microsoft described the ray tracing pipeline from ray shader generation, scheduling, and acceleration structure, all the way to the shading of the actual game. This time, the company has shared insights into areas such as clustered geometry, partitioned top-level acceleration structures (TLAS), and indirect acceleration structure operations.

Firstly, Microsoft introduces the concept of clustered geometry. Readers need to understand that the core graphics elements are triangles, which are the building blocks of the 3D worlds we have today. However, DXR clustered geometry treats groups of nearby triangles as common building blocks or multiple building blocks, allowing the GPU to build, move, and instantiate geometry in bulk. Instead of handling this separately with multiple calls for triangles, the GPU's task is now much more simplified. DXR even defines compact vertex encodings and predefined template formats to ensure that the GPU memory and bandwidth required to execute bulk geometry building and moving are sufficient. As a result, the GPU doesn't have to update or duplicate existing geometry, and DXR will help render foliage, crowds, and in-game props once, allowing them to be moved around easily. This reduces the GPU load and improves the performance of ray tracing in games.

Intel has released its latest 101.8626 WHQL Arc GPU graphics drivers, adding day-one support for Death Stranding 2: On the Beach and Everwind games, as well as introducing the new Intel Graphics Shader Distribution Service, which should improve first load times by up to 2x on Intel Arc B-series and Intel Core Ultra Series 3 and Series 2 CPUs with Intel Arc GPUs. The Graphics Shader Distribution Service is currently limited to a dozen games, so hopefully Intel will extend the list further with the future driver release. In addition, the new Intel Arc GPU 101.8626 WHQL graphics drivers also improve game performance in the Nioh 3 game on Intel Arc B-series GPUs by up to 9 percent at 1080p resolution with Ultra settings.

The new Intel Arc GPU driver release also fixes a couple of issues seen with previous driver releases, including an application crash with ray tracing enabled in the Naraka Bladepoint game, cinematic corruption in Hogwarts Legacy, and visual corruption in the viewport while resizing the window with HDR enabled in Davinci Resolve Studio. These issues are fixed on Intel Arc B-series GPUs and Ultra Series 3 CPUs with Intel Arc GPUs. Since it is a major WHQL release, Intel is listing all new known issues that are left to be fixed with future driver releases.

DOWNLOAD: Intel Arc Graphics Driver 101.8626 WHQL

Samsung Electronics Co., Ltd. today announced it has signed a Memorandum of Understanding (MOU) with AMD to expand their strategic collaboration on next-generation AI memory and computing technologies. The signing ceremony was held at Samsung's most advanced chip manufacturing complex in Pyeongtaek, Korea, attended by Dr. Lisa Su, Chair and CEO of AMD, and Young Hyun Jun, Vice Chairman & CEO of Samsung Electronics.

"Samsung and AMD share a commitment to advancing AI computing, and this agreement reflects the growing scope of our collaboration," said Young Hyun Jun, Vice Chairman & CEO of Samsung Electronics. "From industry-leading HBM4 and next-generation memory architectures to cutting-edge foundry and advanced packaging, Samsung is uniquely positioned to deliver unrivaled turnkey capabilities that support AMD's evolving AI roadmap."

MSI, a global leader in high-performance server solutions, today unveils its latest AI infrastructure portfolio built on NVIDIA's modular architectures, including the NVIDIA MGX platform and NVIDIA DGX Station technology. Designed to accelerate AI training, large-scale inference, HPC, edge, and next-generation data center workloads, MSI's expanded lineup delivers exceptional scalability, performance density, and deployment flexibility.

Scalable AI Infrastructure Built on NVIDIA MGX Architecture
Leveraging the modular design of NVIDIA MGX architecture, MSI has developed a comprehensive portfolio of 4U and 6U liquid-cooled servers supporting NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs. The NVIDIA MGX architecture enables flexible CPU selection, high-capacity memory configurations, and seamless integration of high-speed networking - empowering enterprises to deploy infrastructure tailored to diverse workload requirements, from data center deployments to edge applications.

Strauss Zelnick, the CEO of Take-Two Interactive, has a somewhat complicated relationship with artificial intelligence, having previously expressed an interest in AI NPCs for more natural conversations while also having confirmed that GTA VI will feature no generative AI. Now, in a recent interview with The Game Business, Zelnick has once again commented on the capabilities and applications of generative AI. Addressing a question about Google's recent Project Genie showcases, Zelnick has dismissed the idea that generative AI could be used as a one-stop-shop for game development, saying that the gaming industry has always used technology to create great entertainment," adding that "an advance in technology that allows us to do our work better and quicker is great for us."

Zelnick dismisses the idea that AI projects like Genie are a threat to the gaming industry and to game developers, commenting that "it's quite obvious that creation tools are a benefit to our industry," and that the notion that "AI tools can somehow create big hits kind of doesn't stand to reason." He reasons that generative AI may help developers create game assets, but that creating a hit game requires human engagement and creativity. Zelnick rounds out the AI discussion by emphasizing that Take-Two's goal is to create engaging, entertaining games, and that this requires creativity, adding that "technology can assist with that mission, but technology on its own will not replace the fulfillment of that mission." The executive goes on to explain that "the notion that somehow new tools would allow an individual to push a button and generate a hit and bring it to many millions of consumers around the world, it's a laughable notion." It's worth noting that there have been recent layoffs, like those at EA, that have been attributed to or followed by an increase in AI adoption, indicating that, while AI may not be capable of replacing artists and developers from a technical standpoint, it does not necessarily mean there is no threat.

RPCS3, the multi-platform, open-source PlayStation 3 emulator that recently announced support for 70% of the PlayStation 3 game library, has just added a UX workflow to automatically add emulated games to your Steam library. This effectively works the same way as adding a third-party game to your Steam library from Steam, but it eliminates the extra step of opening Steam and manually adding the game launcher to your Steam library. Games added this way will automatically add the game's PS3 art included with the game files in the emulator, to boot.

It's a small UI change, in the grand scheme of things, but it should help simplify game emulation and make it easier for gamers to play their emulated games via RPCS3. Being able to add third-party games to a game library is a nigh-essential feature that even Microsoft recently added to the Xbox App for Windows. RPCS3 adopting support for seamless Steam library integration could effectively let Linux and SteamOS players go from cold boot to playing a game all from a controller using Steam Big Picture mode—no keyboard or mouse necessary.

Crimson Desert has had no shortage of hype leading up to its March 19 launch date, but research firm, Alinea Insights, has put a number on that hype. Based on Steam data and sales approximations, Alinea estimates that the new open-world action-adventure game has already surpassed $20 million in Steam revenue alone. This is based on an estimated 400,000 pre-orders ahead of launch. Currently, Crimson Desert sits at the top of Steam's Top Sellers charts, which ranks games by revenue earned, having beaten out long-standing chart-topper, Counter-Strike—at least in the US—and the more recent indie hit, Slay the Spire 2, which recently dropped from first to fourth place.

Likewise, according to SteamDB, at the time of writing, Crimson Desert is the fourth most-wishlisted game on Steam, boasting over 170,000 followers on Steam. The excitement leading up to the launch of Crimson Desert has been building for a good long while now, but the game, and its development team, earned a significant boost in its reputation online after a very forgiving set of minimum hardware requirements was announced for Crimson Desert in a time that games seem to be getting more demanding with every passing launch. Alinea attributes much of Crimson Desert's success to the studio's community building efforts and authentic, transparent approach to community engagement, much like the recent gameplay demo showed off on the PlayStation Japan YouTube channel.

At GTC 2026, NVIDIA announced its next-generation DLSS installment, version 5. However, after the community expressed significant backlash over the goals of DLSS 5, NVIDIA CEO Jensen Huang addressed the criticism surrounding the upcoming technology. In a Q&A session, NVIDIA provided Tom's Hardware with a response from a gamer's perspective about the objectives being pursued, as much of the negative feedback stemmed from the perceived visual downgrade in Resident Evil Requiem with the enhancements of DLSS 5. "Well, first of all, they're completely wrong. The reason for that is, as I have explained very carefully, DLSS 5 fuses the controllability of geometry, textures, and every aspect of the game with generative AI." What the CEO of the world's largest company is trying to convey is the controllability offered by DLSS 5 and how it provides developers with an option to enhance visuals.

For example, much of the criticism focuses on how DLSS 5 alters the visual definition of various in-game elements beyond the original developer's intention. Jensen Huang explains that DLSS 5 is not merely a post-processing tool for games but rather a system that gives developers generative control at the geometry level, not just a filter applied on top of the graphics processing pipeline. Huang noted that developers have complete control over what the technology does from the very beginning. DLSS 5 uses the game's color and motion vectors for each frame, and the model aligns with what the game developer originally intended for the image. Even the extent to which DLSS 5 enhances visual fidelity can be controlled, and it is entirely up to the developer to decide. "It's not post-processing, it's not post-processing at the frame level, it's generative control at the geometry level," iterated Huang.

There has been a lot of discussion about Marathon's gameplay, pacing, and potential shortcomings in the weeks since the game launched. One feature that has seemingly been requested all over since the game launched is a duos queue for squads of two players, and it looks as though Bungie is indulging in the fans' requests with an imminent experimental duos queue. According to an X post by Marathon's game director, Bungie will start testing for a duos queue on Wednesday, 18 March, at 10 AM PT. Given that Bungie is even testing a duos queue suggests that there is enough player feedback for the game studio to consider adding the feature to the game at large. There are still questions about how two-player teams will fare in a game that has notoriously high time-to-kill in non-PvP encounters, but that's perhaps one of the gameplay aspects that will need tuning and experimentation to nail down.

This is an experimental test that will be limited to the Perimeter zone, and there are a few minor details to note if you're planning to hop into the new game mode. For starters, and likely the most notable issue players might run into, the duos queue will not have any matchmaking, meaning duos will only be available to players already in a squad with two players. The director is also clear that "some things will be jank," suggesting this will be far from a polished experience for players, but he hopes that any data gathered during the testing can be used to expand the duos queue in the future. Players interested in joining the duos queue can select "Perimeter - Duos" from the zone selection screen, but this will not be the final UX flow, leaving Bungie with a lot of headroom to grow the Marathon duos gameplay experience going forward.

The PS5 release of Starfield has long been rumored for 2026, but it was unclear when exactly the game would launch for Sony's gaming platform, but Bethesda has now confirmed that Starfield will launch on the PS5 in April 2026. Specifically, the launch is slated for an April 7 release. Starfield's PlayStation 5 release will coincide with the launch of a free game content update and a new story DLC, all of which was detailed in a recent developer deep-dive on the Bethesda Softworks YouTube channel.

Bethesda says that the Free Lanes update was heavily guided by player feedback, and that it will be the biggest update since the game's launch. Free Lanes will add a new vehicle and spacesuit, as well as new weapons to find and POIs for players to explore and interact with. It also adds a cruise mode to speed up interplanetary travel and allow players to interact with their crew and ship on longer trips. There are also a new material to enhance weapons and ships, and an expansion to the progression system to give players more to do in space and on the planets they encounter. Enemies have also received an update to make enemy encounters more challenging and varied.

Following the recent early performance benchmarks of the Intel Core Ultra 5 250K Plus, Intel's Arrow Lake-S Core Ultra 3 205T has broken ground on PassMark, and it's giving the Core Ultra 5 255 and Core Ultra 5 255T a run for their money, at least in single-core benchmarks. Multithreaded benchmarks paint something of a grimmer picture for the Core Ultra 3 205T, though—likely due to the reduced core and thread count of the 205T, which features just eight cores and eight threads, likely four P-cores and four E-cores.

The Core Ultra 3 205T scored a respectable 4,561 points in the single-core benchmark, beating out the Core Ultra 5 255 and 255T by 3.2% and 6.6%, respectively. When it comes to multithreaded tasks, however, the Core Ultra 3 205T falls behind. In PassMark's benchmarks, the 205T was nearly 15% behind the Core Ultra 5 225. Still, the Intel Core Ultra 3 205T comes in ahead of the AMD Ryzen AI 5 435 by a significant margin in both single- and multi-core tests, and the Core Ultra 3 205T seems to have been tested at a peak TDP of 35 W, all of which may make it worth considering for lower-end office builds.

The Xbox App has been undergoing an overhaul for some time, and the latest update now allows users to add apps, games, and virtually any third-party software to its library. Windows Central has tested this feature, providing a preview of the process. Although these third-party games and software are not linked to any Microsoft service, the Xbox App offers a centralized location for launching them as shortcuts. For example, Steam has offered a similar feature for years, allowing gamers to add games installed from third-party stores to the Steam Client, but only as shortcuts for launching, not as official sources. This means that any updates to third-party applications will still be managed by their respective apps or clients, with the Xbox App serving merely as a unified shortcut hub.

The process is quite simple. After opening the Xbox App on your PC, go to the "My Library" section, find the small "+" icon in the top right corner, and click it to see suggested additions. If your application isn't listed among the suggested files, the Xbox App enables you to use File Explorer to manually browse for files and locate what you want to display. Nearly any .exe file can be added, including games, productivity apps, or almost anything you can imagine. The only limits are your imagination or something that hasn't been tested yet. For users who want to use the Xbox App as a single app launcher, this feature allows you to embed any game or app within the same user interface, which is a nice option for those who enjoy the Xbox App's user experience.

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.

System76 has refreshed its Thelio Mira desktop series with a focus on better thermals, easier serviceability, and a cleaner overall design, while it's still all built in-house at the company's facility in Denver. The new chassis mixes aluminium, steel, and a tempered glass front panel, with a vertical control bar consolidating the power button and front I/O into one clean strip. System76 uses steel fasteners throughout, and the panels are designed for quick access when you need to get inside for upgrades or maintenance. On the cooling side, System76 made improvements, the company claims up to 19% higher sustained CPU clock speeds and temperature drops of up to 13.5°C thanks to liquid cooling and revised airflow.

System76 also made changes to the specs as it is built around an ASRock X870 Pro RS WiFi motherboard. Processor options top out at the Ryzen 9 9950X or 9950X3D, memory goes up to 192 GB of DDR5, and storage can reach 28 TB spread across NVMe and SATA drives. As for the GPU, you can choose a single graphics card up to an RTX 5090 or Radeon RX 9070 XT over PCIe 5.0. Connectivity includes USB4, 2.5 GbE, and Wi-Fi 7, with a mix of front and rear USB ports. PSU requirements start at 750 W and go up to 1000 W for the beefier GPU configs. As with other System76 products, the Thelio Mira carries open-source firmware elements and is built with long-term usability in mind.

Crimson Desert's official March 19 launch is just around the corner, and in the lead-up to the launch, PlayStation Japan and Pearl Abyss have shown off the upcoming action-adventure game running on base PS5 hardware with respectable image quality and performance. The presenters of the Play! Play! Play! broadcast play through the first few minutes of the game's prologue and show off a few tutorial scenes, and, although it's difficult to gauge image quality directly, there is some information to be gleaned from the broadcast. Overall, performance and image quality seem to align with the promises made by the minimum hardware requirements published earlier this month—even if there are some upscaling artifacts visible in finer textures, like character hair.

The most notable aspect of the gameplay footage is that there are no obvious framerate issues, texture pop-ins, or stuttering visible. Even at longer draw distances, image quality seems to be well controlled, and motion remains smooth, even during high-action scenes where the load would generally increase. The demonstration has been heartening for PS5 players, since there were suspicions that the PS5 gameplay footage was being kept under wraps ahead of the game's launch due to lackluster performance. The broadcast also serves to give players a taste of the challenging combat, puzzle-platformer mechanics, exploration, and atmospheric world they can expect from the game's launch.

Echo Foundry Interactive, an independent game studio focused on building the next generation of music and rhythm experiences and founded by industry veterans behind the Guitar Hero, Rock Band, and DJ Hero franchises, today announced that Sound System, the highly anticipated next-generation rhythm game, will launch on Steam on October 16, 2026 for $24.99.

Sound System revives the rhythm game genre with intense gameplay, built-in creator tools, and a community platform for artists and players to share music-driven creativity. Shred guitar, play bass, or sing vocals using keyboard, controller, compatible guitar controllers, or microphone. Enjoy local split-screen or online modes, including co-op band performances with shared multipliers and effects, and head-to-head competitive modes for stage control battles.

Supermicro, Inc. (NASDAQ: SMCI), a Total IT Solution Provider for AI, Cloud, Storage, and 5G/Edge, today unveiled one of the industry's first context memory (CMX) storage server as part of NVIDIA STX reference architecture announced NVIDIA GTC 2026. STX is a new modular reference architecture from NVIDIA which is designed to accelerate the full lifecycle of AI.

"Supermicro continues to be first to market with new rack scale architectures designed to exceed the needs of a rapidly evolving AI Factory customer base," said Charles Liang, president and CEO of Supermicro. "Building upon last year's introduction of the Petascale JBOF (Just a Bunch of Flash), where we proved the feasibility of a JBOF powered by NVIDIA BlueField-3 DPUs, we have developed the CMX storage server. Our prototype of the latest storage architecture demonstrates the level of our collaboration with NVIDIA, and our commitment to be first-to-market with game changing technologies."

Dell has given its Alienware gaming laptop lineup a boost with Intel's newest Core Ultra 200HX Plus series processors, which are part of the Arrow Lake-HX Refresh. This update comes after Intel rolled out its new mobile chips, including the Core Ultra 9 290HX Plus and Core Ultra 7 270HX Plus. These chips are designed to handle demanding gaming and workstation tasks, and they come with extra features like the Intel Binary Optimization Tool. The updated Dell lineup covers 18-inch and 16-inch laptops from the Alienware Area-51 series and Alienware 16X Aurora. The 18-inch model continues to target maximum performance, now configurable with up to the Core Ultra 9 290HX Plus, a 24-core chip with boost clocks reaching 5.5 GHz, while the Core Ultra 7 270HX Plus offers a 20-core option with boost up to 5.3 GHz.

The 16-inch models bring more notable changes. As previously announced in January at CES 2026, both the Alienware 16 Area-51 and 16X Aurora now feature anti-glare OLED panels, keeping the 2560 × 1600 resolution and 240 Hz refresh rate, but improving response time to 0.2 ms and increasing peak brightness to 620 nits, up from 500 nits on previous LCD configurations. The Alienware 16X Aurora also gets a GPU upgrade, now configurable up to an NVIDIA GeForce RTX 5070 Ti, replacing the previous RTX 5070. Memory support stays the same, with the 16X Aurora maxing out at DDR5-5600, while storage choices go from 1 TB to 4 TB, with PCIe 4.0 support on some setups.

Corsair, a leading maker of performance gaming peripherals, announced the release of the 99% form factor, low-profile VANGUARD AIR 99 WIRELESS Optical-Mechanical Gaming Keyboard. Equipped with low-profile Corsair OPX optical switches, 8,000 Hz hyper-polling, FlashTap SOCD handling, versatile tri-mode connectivity, and an aluminium frame, it's built on a rock-solid foundation of gaming performance.

Premium gasket mounting, five layers of sound dampening, and a brilliant, integrated LCD screen solidify the VANGUARD AIR 99 WIRELESS as a formidable piece of competitive gaming gear that excels in all aspects of daily life. One of our thinnest keyboards ever made, it measures just a scant 18 mm, perfectly designed for modern aesthetic sensibilities. Equipped with Elgato Stream Deck integration and programmable SD-keys, it streamlines daily workflows into single button presses. VANGUARD AIR 99 WIRELESS is an elegant solution that effortlessly delivers high-performance gaming needs and optimizes productivity tasks.

Taiwanese PC case specialist Antec and Austrian quiet cooling expert Noctua have teamed up to create a custom, further refined Noctua Edition of Antec's award-winning Flux Pro chassis. Upgrading the Flux Pro with Noctua's latest NF-A14x25 G2 and NF-A12x25 G2 flagship fans for even better low-noise cooling performance, the new Noctua Edition forms an ideal basis for ultra-quiet, high-end builds.

"Antec has been at the forefront of PC case design for more than two decades, and we're excited to collaborate with such a renowned, iconic manufacturer to introduce the very first Noctua Edition chassis," says Roland Mossig (Noctua CEO). "The Flux Pro has been rightfully praised for its exceptional quiet cooling potential, so it was an obvious candidate for the project. Once we integrated our latest flagship fans and saw how much further we could reduce noise levels while maintaining similar component temperatures, it quickly became clear that this is a worthy Noctua Edition."Editor's Note: Check out the TechPowerUp Review of this case.

Teledyne e2v is pleased to announce the start of full production of its 16 GB DDR4-X1 Flight Model (FM), expanding its portfolio of high-density, radiation-tolerant memory solutions for space applications. The new device is designed to support the growing processing and data storage requirements of AI-enabled satellites, large constellations, broadband Internet-from-Space, Direct-to-Device services, and optical inter-satellite communications. By combining high memory density, radiation resilience, and a compact footprint, the component enables spacecraft to handle increasingly demanding onboard computing workloads.

Initial samples of the 16 GB DDR4-X1 Flight Models were delivered to customers in October 2025, allowing early system integration and evaluation. The device supports data rates up to 2400 MT/s, provides single-event latch-up immunity above 43 MeV·cm²/mg, and offers radiation tolerance up to 35 krad TID, enabling reliable operation in mission-critical space environments.

Industry-leading embedded solutions provider AAEON (Stock Code: 6579) has announced the release of the ATX-Q870A, an ATX industrial motherboard supporting Intel Core Ultra Series 2 Processors (formerly Arrow Lake-S) and up to 256 GB of dual-channel DDR5-6400 system memory. Designed to accommodate up to 125 W CPUs from the new Intel CPU platform, the ATX-Q870A can leverage up to 24 cores and 24 threads of computing power alongside up to 36 TOPS of AI performance through the series' integrated CPU, GPU, and NPU. As such, AAEON has positioned the product as a foundation upon which compute-intensive applications such as industrial automation systems and high-performance workstations can be built.

Despite its impressive processing capabilities, it is the expandability offered by the ATX-Q870A that is most likely to catch attention. Of particular note are the board's two PCIe Gen 5 and five PCIe Gen 4 slots, which allow it to support high-spec GPUs and AI accelerators, NVMe storage, and task-specific modules like serial cards, sensor interfaces, and low-speed NICs simultaneously.

ASUS Republic of Gamers (ROG) is proud to launch the 2026 ROG Strix G16 and Strix G18, two incredible gaming laptops featuring the latest hardware and technology to deliver incredible performance to gamers everywhere. Boasting the all-new Intel Core Ultra 9 290HX Plus processor and up to an NVIDIA GeForce RTX 5080 Laptop GPU, these machines are built from the ground up for enthusiast gamers. Both the ROG Strix G16 and G18 feature the latest ROG displays, cooling, and tool-less chassis designs that allow gamers to seamlessly upgrade critical components. Later this year, the flagship ROG Strix SCAR 18 will also be unveiled, delivering flagship performance in an incredibly sleek ROG chassis.

Elite performance for every arena
The 2026 Strix G16 and G18 gaming laptops are engineered for players and creators who demand uncompromising performance across competitive esports titles, visually rich AAA games, and intensive content‑creation workflows. Both models feature the latest Intel Core Ultra 9 290HX Plus processor, delivering exceptional multi‑threaded power and next‑generation AI capabilities.

Intel today announced the launch of its new Intel Core Ultra 200HX Plus series mobile processors, giving gamers and professionals new high-performance options in the Core Ultra 200 series family. Optimized for advanced gaming, streaming, content creation, and workstation use, the Intel Core Ultra 200HX Plus series introduces two new processors - Intel Core Ultra 9 290HX Plus and Intel Core Ultra 7 270HX Plus. These processors add new features and architectural refinements, including support for the new Intel Binary Optimization Tool, a first-of-its-kind binary translation layer optimization capability that can improve native performance in select games.

"With the introduction of the Intel Core Ultra 200HX Plus series, we're pushing mobile computing performance even further for the gamers, creators, and professionals who demand the best. With higher die-to-die frequencies and our new Intel Binary Optimization Tool, the new Intel Core Ultra 9 290HX Plus and Ultra 7 270HX Plus deliver meaningful, real‑world performance gains so users can experience smoother gameplay, faster creation workflows, and more responsive workstation performance", Josh Newman, General Manager and Vice President of Product Marketing, Client Computing Group.

Silicon Motion Technology Corporation, a global leader in designing and marketing NAND flash controllers for solid-state storage devices, today announced that it will showcase a rich portfolio of differentiated enterprise SSD controllers and PCIe NVMe BGA boot SSD at NVIDIA GTC 2026, Booth #3015, purpose-built to meet the evolving requirements of the NVIDIA AI ecosystem. As AI models scale, inference architectures are extending beyond HBM and System DRAM into high-performance NAND storage tiers, as reflected in NVIDIA's ICMS initiative. In this new architecture, NAND-based Storage becomes a performance-critical layer that requires deterministic latency and quality-of-service differentiation.

Silicon Motion delivers a vertically integrated storage solution encompassing advanced SSD controller design, full firmware development, and compact Reference Design Kits (RDKs) aligned with leading enterprise form factors such as E1.S, E1.L, E3.S, E3.L, and U.2 for AI server deployments. The company also provides enterprise PCIe NVMe BGA boot SSD with strong endurance and long-term operational stability, deployed in AI systems.

Logitech G today announced the long-anticipated RS H-Shifter, the latest addition to the renowned Racing Series Ecosystem. Designed to deliver unmatched realism, tactile control, and game-changing durability, this advanced 7+R manual shifter is tailor-made for anyone passionate about authentic racing experiences. For years, racers have yearned for a product that combines modern engineering with the timeless precision of manual control, and Logitech G has delivered. From gripping rally runs in Assetto Corsa Rally to flawless drifting in Forza Horizon, the RS H-Shifter gives racers the tactile realism and precise control needed to dominate the virtual track.

"There's a strong demand from car enthusiasts worldwide for the connection and control that a manual shifter offers," said Richard Neville, Head of Product, Simulation at Logitech G. "The RS H-Shifter's engaging, racing gearbox feel, is engineered to reliably deliver the elevated experience that is expected from Racing Series and PRO products."

Yesterday, NVIDIA unveiled its latest DLSS 5 technology, offering gamers the first real-time neural rendering. However, even after the announcement, many questions arose about what DLSS 5 is capable of and how it will work with games. NVIDIA released a FAQ to address some common inquiries. The primary goal of DLSS 5 is to enhance visual fidelity through various techniques that create scenes with photorealistic lighting and materials. Perhaps the most interesting aspect is that DLSS 5 will honor the original artistic intent by using the game's color and motion vectors for each frame, anchoring the DLSS 5 model to the specific setting. This keeps the output in line with what the game developers originally envisioned for each frame. DLSS 5 will add visual enhancements that help each frame undergo an overhaul.

This overhaul is completed in several steps. The first is cinematic lighting, achieved through complex effect reconstruction for realistic skin glow, shadows, and more. Next is material depth—DLSS 5 applies micro-realism to the surface of any object, such as a rock or a wall, delivering a realistic texture that enhances the game. NVIDIA highlights that its latest DLSS installment offers temporal consistency, meaning the image quality is fine-tuned frame by frame to closely follow the game content, ensuring visual enhancements remain consistent. Interestingly, this technology will work alongside existing techniques like path tracing, where path tracing provides lighting accuracy, and DLSS enhances lighting photorealism. This means path tracing improves overall shadow behavior and reflections, while DLSS 5 makes them as realistic as possible.

Acer today announced refreshed Predator Helios Neo series gaming laptops equipped with the latest Intel Core Ultra 200HX Plus series processors and up to an NVIDIA GeForce RTX 5080 Laptop GPU, bringing the latest performance capabilities to gaming enthusiasts across a range of form factors and display options. The new Intel Core Ultra 200HX Plus series processors power a new class of gaming laptops with significant performance gains over the previous generation. These devices are built for split-second responsiveness, rock-solid FPS, smart tuning, and battery life that keeps players locked in across an extensive selection of games and apps.

Powered by NVIDIA Blackwell, NVIDIA GeForce RTX 50 Series Laptop GPUs bring game-changing capabilities to gamers and creators. Equipped with a massive level of AI horsepower, the RTX 50 Series enables new experiences and next-level graphics fidelity. Multiply performance with NVIDIA DLSS 4.5 and generate images at unprecedented speed.

Giga Computing, a subsidiary of GIGABYTE and a leader in accelerated computing and infrastructure solutions, today announced new enterprise AI solutions that support NVIDIA Vera CPU and NVIDIA Vera Rubin platform, well as a new AI factory in Taiwan. The GIGABYTE booth at NVIDIA GTC shows scalable AI solutions that not only focus on performance and efficiency but also incorporate the software and infrastructure needed to build AI factories and other large GPU clusters. The GIGABYTE booth staff is ready to introduce the latest hardware and software for success in deploying accelerated computing.

Personal AI Supercomputers
With products spanning all segments of the supercomputer space, Giga Computing showcases professional desktop and deskside solutions that are ideal for AI development and accelerating AI training and inference workloads. These solutions are being used by data scientists and researchers in research institutions, government agencies, and enterprises.

As the era of pervasive AI reshapes industries worldwide, ASRock Industrial today announced the AI BOX-A395, a compact yet powerful AI box that brings the performance of an ultimate AI workstation into a single system. Powered by AMD Ryzen AI Max+ 395 processors, delivering up to 50 TOPS of AI acceleration while integrating CPU, GPU, and NPU within a compact system. With support for up to 128 GB LPDDR5x-8000 unified memory, it enables large AI models and data-intensive workloads to run directly on-device, delivering responsive AI processing while reducing reliance on cloud infrastructure.

Designed for enterprises, developers, and system integrators, the AI BOX-A395 supports the AI everywhere ecosystem by translating large-scale AI capabilities into practical local deployment. By combining high compute density, integrated AI acceleration, and rich I/O connectivity, the system provides a scalable foundation for applications ranging from AI model and application development, engineering and 3D design, and high-resolution content creation and media production.

ASUS is proud to announce the launch of the 2026 TUF Gaming A16, F16, and A18 gaming laptops. The TUF Gaming A16 and F16 boast two impressive display options, either a gorgeous 2.5K 120 Hz OLED panel or an ultra-fast 2.5K 300 Hz IPS display. Both models feature advanced anti-reflection coatings on the panel for increased immersion, while the TUF Gaming F16 also comes equipped with the all-new Intel Core Ultra 9 processor 290HX Plus. The larger TUF Gaming A18 sports up to an NVIDIA GeForce RTX 5070 Ti Laptop GPU and a lightning-fast 2.5K 300 Hz IPS display for a truly impressive and immersive gaming experience.

16-inch upgrades
The 16‑inch TUF Gaming A16 and F16 receive major upgrades this generation, headlined by two premium new display configurations designed to elevate both immersion and competitive play. Users can choose between a stunning 2.5K 120 Hz OLED panel featuring a Corning DXC advanced anti‑reflection coating, or an ultrafast 2.5K 300 Hz IPS panel enhanced with ACR anti‑reflection technology for clearer visibility in bright environments. These advanced coatings significantly reduce glare while preserving color accuracy and contrast even when viewed at off angles, ensuring gameplay remains sharp and distraction‑free and increasing immersion across your entire gaming library.

Jeff Kaplan, formerly a Blizzard executive, recently announced both a new studio and a new game, The Legend of California, which is slated to launch in 2026. Since the game's launch, the reception has been somewhat mixed, but that seemingly hasn't dampened the spirits of the studio or its executive. In a recent 10-hour livestream of The Legend of California, Kaplan addressed some of the negative comments he'd seen online, mainly taking aim at players who were criticizing the game, but who he seems to think weren't the target market in the first place.

The specific demographic that spurred Kaplan's comments were Overwatch players who have had what he describes as a "nerd rage-out," and expressing their frustration at Kaplan's chosen genre, visuals, or game design for The Legend of California. In response to these players, Kaplan said "if a game comes out, and you don't want to play it, and you've never played it, shit the f**k up---no one cares. We don't need to hear that you aren't into it." He goes on to question the reasoning behind players voicing their opinions on a video game they're never going to play, asking "Who cares about my opinion if I'm not going to play it, and if I've never played it? Why does my opinion matter on that?"

MSI today announced the launch of XpertStation WS300 on NVIDIA DGX Station Architecture, a next-generation deskside AI supercomputer built to support the accelerating demands of large language models (LLMs), generative AI, and advanced data science workflows. Powered by NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, supporting up to 748 GB of large coherent memory and dual 400GbE networking, the platform extends advanced AI infrastructure capabilities into a compact deskside deployment model and is available for order starting today.

"MSI has a strategic vision to advance AI-first computing," said Danny Hsu, General Manager of MSI's Enterprise Platform Solutions. "With NVIDIA, we are defining the next era of AI infrastructure, bridging centralized performance and distributed innovation, and enabling organizations to move from experimentation to production with greater speed, scale, and confidence."

HPE (NYSE: HPE) today announced significant innovations to the NVIDIA AI Computing by HPE portfolio focused on large-scale AI factories and supercomputers that enable customers to scale, deploy efficiently, and gain faster time-to-insight. The full-stack AI solutions with NVIDIA include tightly integrated compute, GPUs, networking, liquid cooling, software, and services designed for at-scale and sovereign environments. AI-forward organizations and leading research institutions, including Argonne National Laboratory, HLRS, Hudson River Trading (HRT), and the Korea Institute of Science and Technology Information (KISTI), have chosen HPE AI infrastructure and AI factories with NVIDIA to advance innovation.

HPE brings NVIDIA AI solutions to its industry-leading supercomputing platform
Research laboratories, sovereign entities and large enterprises are rapidly adopting AI to enhance traditional high performance computing (HPC) workloads. For organizations seeking to significantly expedite scientific discovery, HPE is making the following NVIDIA products available on its second-generation exascale-class supercomputing platform designed to unify AI and HPC - the HPE Cray Supercomputing GX5000.

ASUS today unveiled its fully liquid-cooled AI infrastructure at NVIDIA GTC 2026 (Booth# 421), delivering a comprehensive, end-to-end solution powered by the NVIDIA Vera Rubin platform. Under the theme Trusted AI, Total Flexibility, this customizable framework—from rack-scale AI Factories, desktop AI supercomputing, Edge AI to Enterprise AI solutions—enables enterprises and cloud providers to build high-performance, energy-efficient large-scale AI clusters with unmatched efficiency and dramatically reduced PUE and TCO.

As a provider of NVIDIA GB300 NVL72 and NVIDIA HGX B300 systems, the flagship ASUS offering is the ASUS AI POD built on the NVIDIA Vera Rubin platform—a liquid-cooled, rack-scale powerhouse designed for massive AI workloads. Through strategic partnerships with leading cooling and component providers, ASUS offers diverse cooling modalities, tailored thermal solutions, and redundancy to meet any enterprise requirement. Proven by global client successes, ASUS provides expert consultation, a broad portfolio of AI and storage solutions, seamless infrastructure deployment, application integration, and ongoing services—combining scalability, and sustainability to drive business value and intelligence.

At NVIDIA GTC 2026, Micron announced it is in high-volume production of three products at once, all timed around and designed for the Vera Rubin platform. The headline is HBM4 with the 36 GB 12-high stack started shipping in volume in Q1 2026, built for NVIDIA Vera Rubin. It hits over 11 Gb/s pin speeds for more than 2.8 TB/s of bandwidth, 2.3 times what HBM3E offered, while also improving power efficiency by over 20%. Micron has also shipped early samples of a 16-high 48 GB variant to customers, a 33% capacity bump per HBM placement over the 12H stack. Micron also announced that its 192 GB SOCAMM2 memory modules are now in high-volume production. Designed for Vera Rubin NVL72 systems and standalone Vera CPU platforms, it enables up to 2 TB of memory and 1.2 TB/s of bandwidth per CPU, with the broader SOCAMM2 portfolio spanning from 48 GB to 256 GB.

On the storage side, Micron 9650, the industry first PCIe Gen 6 data center SSD designed specifically for NVIDIA BlueField-4 STX architecture, is in mass production as we already reported here. It boasts sequential read speeds of up to 28 GB/s and can handle 5.5 million random read IOPS, essentially doubling the read performance of its Gen 5 predecessor. Furthermore, it offers a performance-per-watt ratio that is twice as efficient.

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.

Phison Electronics (8299TT), a global leader in NAND flash controllers and storage solutions, today announced its GTC showcase at booth 119, demonstrating how multi-tier memory architecture supports larger models and long-context inference on NVIDIA-powered local AI platforms. The industry is facing a growing memory constraint while demand for AI-ready platforms continues to surge. Fine-tuning and inference on proprietary data require massive compute and memory resources, creating investment challenges for organizations. These rising solution costs and workflow bottlenecks are slowing time-to-market for revenue-generating innovation. To address this challenge, Phison introduced aiDAPTIV technology for local and edge AI use cases. By leveraging Pascari SSDs as a new AI memory tier, aiDAPTIV technology intelligently extends and manages AI working memory across GPU memory, system RAM and flash.

Today's announcement showcases how aiDAPTIV applies these multi-tier memory architecture principles to local AI systems as NVIDIA AI infrastructure advances GPU memory capabilities to support inference workloads in data center environments. Built on high-endurance flash optimized for sustained paging and context retention, aiDAPTIV supports memory-intensive inference and fine-tuning workloads under fixed hardware configurations. The aiDAPTIV flash-based memory tier enables organizations to support these evolving workloads on local systems while maintaining data privacy and improving long-term infrastructure efficiency.

Lenovo introduces the next-generation of AI workstations optimized for on-device AI development, inference and creation, including the ThinkPad P14s i Gen 7, ThinkPad P14s Gen 7 AMD, ThinkPad P16s i Gen 5, ThinkPad P16s Gen 5 AMD, ThinkPad P1 Gen 9 and the powerhouse desktop ThinkStation P5 Gen 2.

Designed for students, engineers, data scientists and everyone in between, the new workstations pack unprecedented performance capable of tackling even the most demanding workflows, including CAD, BIM, data science, AI development and more. Part of Lenovo's new Hybrid AI Advantage solutions with NVIDIA, the systems accelerate AI adoption, boost business productivity, and bring faster ROI from AI deployments, reflecting Lenovo's commitment to deliver smarter, more adaptive solutions to meet modern challenges.

Samsung Electronics, a global leader in advanced semiconductor technology, today announced the comprehensive AI computing technologies it will showcase at NVIDIA GTC 2026 in San Jose, California, scheduled for March 16-19. As the industry's only semiconductor company offering a total AI solution spanning memory, logic, foundry and advanced packaging, Samsung will exhibit its full suite of products and solutions that enable customers to design and build groundbreaking AI systems. To learn more about Samsung's AI solutions, please visit the company's GTC 2026 booth (#1207).

The centerpiece of Samsung's showcase at NVIDIA GTC 2026 will be the new sixth-generation HBM4, which is now in mass production and is designed for the NVIDIA Vera Rubin platform. Samsung's HBM4 is expected to help accelerate the development of future AI applications, delivering consistent processing speeds of 11.7 gigabits-per-second (Gbps), which exceeds the industry standard of 8 Gbps, and can be enhanced to 13 Gbps.

Penguin Solutions, Inc. (Nasdaq: PENG), the AI factory platform company, today announced the industry's first production-ready KV cache server that utilizes CXL memory technology to address the critical "memory wall" challenge in AI inferencing—Penguin Solutions MemoryAI KV cache server. This innovative solution delivers up to 11 TB of CXL-based memory engineered to optimize performance of enterprise scale inference, including agentic AI. The result is lower latency, higher throughput, increased efficiency of GPU clusters, consistent achievement of stringent service-level agreements (SLAs), and faster time-to-first-token (TTFT).

While model training and tuning is primarily compute-bound and occurs episodically, the continuous memory-bound and latency-sensitive inference workloads required for inference and agentic AI are complex and fundamentally different. Inference demands are typically 30% compute driven (GPU) and 70% memory driven (RAM), elevating the need for greater memory capacity and causing performance bottlenecks and GPU idle time. Accelerating memory-dependent AI processes, Penguin's MemoryAI KV cache server increases memory capacity by integrating 3 TB of DDR5 main memory and up to eight 1 TB CXL Add-in Cards (AICs).