INNO3D proudly announces the launch of its latest enterprise solution, the INNO3D NVIDIA MGX 4U GPU Server with NVIDIA ConnectX 8 SuperNIC - AGS-4UMGX-R1, engineered to empower businesses and institutions with cutting-edge performance for the most demanding workloads. The INNO3D NVIDIA MGX 4U GPU Server leverages the parallel processing power of NVIDIA GPUs to accelerate capabilities across High-Performance Computing (HPC), data analysis, genomics, 3D modeling and imaging, machine learning, simulations, deep learning, and AI training. Designed for professionals, this server is ideal for Large Language Model (LLM) training, cloud AI deployments, and enterprise-scale innovation.

Corsair (Nasdaq: CRSR) today introduced Corsair Pro, a new portfolio of AI systems built for organizations deploying and scaling modern AI workloads. Designed to support multiple workloads from model development and fine-tuning to inference and production deployment, Corsair Pro brings together purpose-built workstations and servers in a single platform family.

With AI teams facing very different infrastructure requirements depending on workload, model size, and deployment stage, Corsair Pro is built to offer a more practical approach. Some users need a workstation for development, experimentation, validation, and local inference. Others need server-class infrastructure for denser compute, shared environments, and production-scale deployment. Corsair Pro is designed to support both.

AMD has formally launched the EPYC 8005 series, codenamed "Sorano" after its announcement at MWC 2026 back in February, and now the full details are in. The new lineup succeeds the EPYC 8004 "Siena" family and sits between the high-end EPYC 9005 and the entry-level EPYC 4005, targeting single-socket low-power platforms with eight to 84 Zen 5 cores across a 95 W to 225 W TDP range. The flagship is the EPYC 8635P featuring 84 cores, 168 threads, 384 MB of L3 cache, and a 225 W TDP with a 1.6 GHz base clock. Compared to its 64-core EPYC 8004 predecessor, AMD claims 40% higher integer performance and 9.5% better performance per watt. Against Intel's 40-core Xeon 6716P-B at 235 W, it offers 2.1 times the core count (84 vs 40), a 10 W lower TDP (225 W vs 235 W), and 91% higher integer performance. AMD also claims 48% better integer performance per CPU watt per dollar versus Intel 72-core Xeon 6776P-B in a single-socket configuration.

The full lineup spans seven SKUs, from the 8-core EPYC 8025P at 95 W up to the 84-core 8635P at 225 W. All models are single-socket only and support six channels of DDR5-6400 ECC memory with up to 3 TB total capacity, 96 lanes of PCIe Gen 5, and full AVX-512 support. AMD is positioning the 8005 series around dense edge deployments and software-defined storage workloads. In those scenarios, core density, I/O bandwidth and power efficiency matter more than raw single-threaded performance.

As artificial intelligence (AI) and data-intensive applications continue to drive demand for real-time processing and low-latency computing, high-performance computing (HPC) is rapidly expanding beyond traditional data centers to the edge. Addressing this growing trend, Avalue Technology has launched the HPS-GNRU1A 1U high-density server system and the HPM-GNRUP industrial server board, combining the latest Intel Xeon 6 platform, PCIe Gen 5 architecture, and high-density NVMe storage to deliver flexible and high-performance solutions for emerging Edge HPC applications.

Unlike conventional HPC platforms primarily designed for centralized data centers, Avalue's latest Edge HPC portfolio is optimized for on-site computing environments. In smart manufacturing, AOI inspection systems require real-time image processing to improve production yield and process stability. In healthcare, image-guided therapy depends on precise and low-latency image analysis, while smart city applications such as airport security screening demand rapid processing of high-frequency imaging data. In life sciences, next-generation sequencing (NGS) workloads also require intensive computing performance for large-scale data analysis. These applications share common challenges, including massive data throughput, real-time processing requirements, and deployment in space-constrained environments, making Edge HPC a critical market focus.

KIOXIA America, Inc. today announced a breakthrough in high-density storage infrastructure, collaborating with Dell Technologies to deliver a 2U server configuration capable of scaling to an unprecedented 9.8 petabytes (PB) of flash storage. By combining the Dell PowerEdge R7725xd server with AMD EPYC processors and 40 KIOXIA LC9 Series E3.L form factor 245.76 terabyte (TB) NVMe SSDs, the companies are enabling a new class of storage-optimized platforms built to meet the demands of AI, large-scale data lakes, and data-intensive enterprise workloads.

Together, Dell Technologies and Kioxia have a long history of enabling scalable infrastructure for data-driven applications. This latest milestone underscores a shared commitment to advancing high-density architectures that improve performance, reduce energy consumption, and maximize data center efficiency.

Micron Technology, Inc. (Nasdaq: MU), today announced it has sampled 256 GB DDR5 registered dual in-line memory modules (RDIMM) to key server ecosystem enablers. The module is built on the company's leading-edge 1-gamma technology, which is capable of speeds up to 9,200 megatransfers per second (MT/s), greater than 40% faster than modules in volume production today. Micron's module employs advanced packaging techniques, 3D stacking (3DS) multiple memory dies connected by through-silicon vias (TSVs). Combined with Micron's 1-gamma DRAM, these innovations provide the capacity, speed and power efficiency required to scale next-generation AI systems. A single 256 GB module can reduce operating power by more than 40% versus two 128 GB modules, enabling greater efficiency for modern AI data centers.

Ecosystem partner validation
Micron is collaborating with key ecosystem enablers to validate the 256 GB 1-gamma DDR5 RDIMM across their respective current and next-generation server platforms. This co-validation ensures broad platform compatibility and accelerates the path to production deployment for data center customers building AI and HPC infrastructure at scale.

HPE (NYSE: HPE) today announced the availability of HPE Compute Scale-up Server 3250, a purpose-built server for in-memory databases, engineered to deliver scalability, resilience, performance, and security for critical business transactional and analytics workloads. Powered by Intel Xeon 6 processors, HPE Compute Scale-up Server 3250 delivers the highest memory available as the first scale-up server validated by the SAP BW Edition HANA benchmark with at least 48 terabytes (TB) of memory.

"HPE Compute Scale-up Server 3250 raises the bar for business-critical infrastructure," said Krista Satterthwaite, senior vice president and general manager, Compute at HPE. "Designed for complex business applications like enterprise resource planning and customer relationship management systems, this system helps customers in ways competitive offerings cannot - allowing them to run faster transactions, gain insights in real time, and support analytics without slowing down business. With simpler operations, always-on reliability, and chip-to-cloud security - we're redefining what scale looks like for workloads with zero-tolerance downtime."

AMD's CPU offerings will soon expand into more categories, specializing in different tasks and use cases. During the latest Q1 earnings call, AMD CEO Dr. Lisa Su confirmed that customers are increasingly interested in specialized EPYC solutions and that future CPU generations will cater to different customer needs. This means that agentic AI will receive a dedicated EPYC CPU SKU, HPC workloads will have their own EPYC CPU SKU, AI training and inference will get a separate EPYC SKU, and cloud workloads will have a distinct SKU. This expansion of task-specialized CPU designs will go beyond the "Venice" CPU platform, which is expected to be released this year. At CES 2026, AMD confirmed that EPYC "Venice" will be a highly dense CPU package with up to 256 cores and 512 threads of "Zen 6c" cores, while the regular "Zen 6" configuration will have a maximum of 96 cores and 192 threads.

With this generation of EPYC CPUs based on the "Zen 6" core IP, AMD will promote "Venice" CPUs broadly, while the "Venice" CPU generation will be complemented by the "Verano" CPUs. These appear to use the same "Zen 6" microarchitecture but have optimizations specifically designed for AI infrastructure. Contrary to previous assumptions that "Verano" would be a "Zen 7" IP, it is now confirmed to be part of the 6th-generation EPYC CPU family. This indicates that AMD is starting its CPU customization for different workloads as early as this year. As the server CPU market is projected to reach $120 billion by the end of this decade, less than 3.5 years away in 2030, it makes sense to have different CPU SKUs for each task and capture a significant portion of that revenue. Server CPUs are also expected to grow at a 35% compound annual growth rate, which is impressive for a relatively mature market.

As data sovereignty and compute performance become strategic differentiators for enterprises adopting AI, the demand for private, on-premises AI infrastructure continues to grow. In response to this shift, QNAP Systems, Inc., a leading innovator in computing, networking, and storage solutions, today introduced the QAI-h1290FX, a next-generation Edge AI storage server designed to empower private deployment of large language models (LLMs), Retrieval-Augmented Generation (RAG) search engines, and generative AI applications.

Built with server-grade AMD EPYC processing, with support for NVIDIA RTX GPU acceleration, and twelve U.2 NVMe/SATA SSD slots, the QAI-h1290FX delivers a high-performance, on-prem AI infrastructure for organizations that demand low-latency inference, full data privacy, and operational control—without relying on the cloud.

Super Micro Computer, Inc. , an AI, Enterprise, Storage, 5G/Edge Total Solution provider, today expanded its Data Center Building Block Solutions (DCBBS) portfolio with new Arm -based server platforms powered by the new Arm AGI CPU, and new Open Compute Project (OCP) ORv3-compliant rack offerings. Supermicro leads the industry with over 20 OCP Inspired systems, which incorporate various OCP technologies and form factors, to simplify open data center deployments.

"Supermicro continues to advance its DCBBS with an expanded portfolio of Arm-based platforms and OCP systems for next-gen AI and HPC," said Charles Liang, president and CEO of Supermicro. "With high-density, liquid-cooled systems and energy-efficient Arm architectures, we enable scalable, flexible data centers that maximize performance-per-watt and accelerate AI adoption across cloud and enterprise environments."

One of the often-touted benefits of console gaming is how physical media enables access to games, even when there is no internet connection for the DRM to verify the game's license. Now, it seems as though the Sony's PS5 will start locking players out of digital-only games if the console has been unable to connect to PSN servers within the last 30 days. This is according to testing done by Hikikomori Media, who found the issue with Wild Arms 4 and Vampire Crawlers, both of which were purchased as digital-only titles in April 2026.

The content creator simulated an offline PS5 by disconnecting it from the internet, resetting the internal clock by removing the CMOS battery, and then booting up the PS5 without an internet connection. Effectively, he tricked the PS5 into thinking that it had been offline for an extended period of time. The end result is that both Wild Arms 4 and Vampire Crawlers were unable to launch after the reboot. It's notable that games purchased before March 2026 would still launch without a hitch. Based on this testing and further testing with a PS4, which revealed a timer showing how long it had been since the console could call home, it seems as though games purchased after March 2026 will need to connect to PSN servers at least once every 30 days in order to be playable.

Rambus Inc. (NASDAQ: RMBS), a premier chip and silicon IP provider making data faster and safer, today announced a SOCAMM2 (Small Outline Compression Attached Memory Module) chipset designed to enable low-power, high-performance LPDDR5X-based memory modules for AI server platforms. The SOCAMM2 chipset represents the first step in a broader Rambus roadmap of LPDDR-based server module solutions, reflecting the company's ongoing collaboration with industry partners to support new memory architectures optimized for evolving workloads in AI data center infrastructure. This new product family extends the comprehensive Rambus offering of complete memory interface chipsets for all JEDEC-standard DDR5 and LPDDR5 memory modules.

The rapid diversification and expansion of data center workloads driven by AI is reshaping system requirements, heightening the need for purpose-built solutions that optimize for power, efficiency, form factor and memory scalability. SOCAMM2 memory modules, based on LPDDR technology, are emerging as an innovative architectural approach to address these challenges by delivering high performance with lower power consumption in a modular, serviceable, board area efficient form factor.

EK by LM TEK is proud to announce the EK-Pro GPU WB H200 NVL, a high-performance full-cover GPU water block designed for the NVIDIA H200 NVL. Engineered for high-density AI server deployments and data center environments, the block features a single-slot form factor optimized for maximum compute density and system compatibility. By covering the entire PCB, the EK-Pro water block directly cools the GPU, HBM memory, and VRM ensuring optimal thermal performance, stability, and sustained performance under demanding workloads. The low-profile, rack-style terminal design reduces overall assembly height, improving chassis compatibility across standard Full Height Full Width (FHFW) server platforms. This enables seamless integration as a drop-in replacement for both active and passive dual-slot air coolers.

Dual-Pass Cooling Engine for Maximum Thermal Efficiency
At the core of the EK-Pro GPU WB H200 NVL is a dual-pass cooling engine, engineered to maintain high flow rate. The coolant flows sequentially through two optimized microfin stacks, cooling each half of the GPU die and HBM memory to ensure uniform heat dissipation. This design maintains consistent flow distribution and thermal performance, minimizing losses even in reversed flow scenarios, while delivering efficient cooling across the entire GPU surface.

EK by LM TEK is proud to announce the EK-Pro GPU Water Block for NVIDIA RTX PRO 6000 Blackwell Server Edition & MAX-Q Workstation Edition GPUs, a high-performance single-slot solution engineered for high-density AI server deployments and professional workstation applications. Designed for NVIDIA RTX PRO 6000 GPUs, this full-cover EK-Pro block actively cools the GPU core, VRAM, and VRM, ensuring optimal performance, stability, and efficient heat dissipation.

The single-slot form factor maximizes GPU density, making it ideal for AI and server infrastructure. Integrated quick-disconnect fittings enable efficient maintenance without the need for disassembly or loop draining, reducing downtime and supporting scalable data center operations. The EK-Pro GPU Water Block for NVIDIA RTX PRO 6000 Blackwell Server Edition and MAX-Q Workstation Edition is available through the EK Shop via our enterprise team. Please note this EK is not compatible with the reference RTX Pro 6000 Workstation Edition GPU.

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.

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."

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.

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.

Super Micro Computer, Inc. (NASDAQ: SMCI), a Total IT Solution Provider for Cloud Computing, AI/ML, Storage, and 5G/Edge, today unveiled its upcoming system portfolio powered by the NVIDIA Vera Rubin platform. As data centers transform into AI factories, producing intelligence at massive scale, agentic reasoning, long-context AI, and Mixture-of-Experts (MoE) workloads are driving demand for an entirely new class of compute and storage infrastructure. Supermicro's NVIDIA Vera Rubin NVL72, NVIDIA HGX Rubin NVL8, NVIDIA Vera CPU systems are being designed and built with Supermicro's Data Center Building Block Solutions (DCBBS) advanced liquid-cooling technology stack to accelerate time-to-market for customers.

"We are entering a new era where every organization requires an AI factory to win in the marketplace, as the demand for inference workloads is reshaping what data center infrastructure must deliver," said Charles Liang, president and CEO of Supermicro. "Supermicro's DCBBS technology stack is being engineered to empower upcoming NVIDIA Vera Rubin NVL72, HGX Rubin NVL8, and Vera CPU systems to give our customers a fast, clear path to deploying next-generation AI factories, at scale. We are excited to provide an early look at these solutions as a testament to being first to market with the infrastructure that will power the next frontier of AI."

SilverStone Technology today announced the RM100, a 1U rackmount server chassis designed for space-constrained server environments. Despite its compact 1U form factor, the RM100 supports up to ATX motherboards, providing system integrators and enterprise users with a flexible and efficient platform for building rackmount systems.

The RM100 is engineered with system configuration flexibility in mind. Its reversible chassis design allows users to choose whether the I/O ports face the front or the rear of the rack, making it easier to adapt to different rack layouts and installation environments. The power supply position can also be adjusted to suit various deployment requirements. In addition, components such as the rail kit, handles, drive bays, and power module are designed to be reversible, offering greater convenience and adaptability during system integration and installation.

To overcome the thermal barriers limiting AI performance, Frore Systems today unveiled LiquidJet Nexus, a new integrated liquid cooling system designed for the AI era. As AI compute densities surge, heat is emerging as a primary constraint, elevating what experts are calling the "AI Thermal Stack" - the infrastructure responsible for extracting heat from advanced computing systems and rejecting it into the atmosphere. The AI Thermal Stack is increasingly determining the compute density, system efficiency, and overall performance of AI Data Centers.

LiquidJet Nexus is designed for liquid cooling ½ U compute trays like NVIDIA Kyber, directly cooling GPUs, CPUs, DPUs, NICs, DC-DC power converters, and VRs. It integrates multiple multi-stage 3D short-loop jetchannel liquid cooling LiquidJet coldplates into a unified system. LiquidJet Nexus delivers 75% higher heat transfer efficiency, keeps components 8°C cooler, eliminating hoses, connectors, and manifolds, and reducing thermal stack weight by 65%. This unified system addresses the thermal demands of next-generation AI hardware.

Game development teams are working across larger worlds, more complex pipelines and more distributed teams than ever. At the same time, many studios still rely on fixed, desk-bound GPU hardware for critical production work. At the Game Developers Conference (GDC) this week in San Francisco, NVIDIA is showcasing a new approach to bring together disparate workflows using virtualized game development on NVIDIA RTX PRO Servers, powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs and NVIDIA vGPU software.

With the RTX PRO Server, studios can centralize and virtualize core workflows across creative, engineering, AI research and quality assurance (QA) - all on shared GPU infrastructure in the data center. This enables teams to maintain the responsiveness and visual fidelity they expect from workstation-class systems while improving infrastructure utilization, scalability, data security and operational consistency across teams and locations.

Advantech (2395.TW), a global leader in industrial IoT, today announced an expanded collaboration with Qualcomm Technologies, Inc. to accelerate the deployment of generative AI (GenAI) at the edge. This collaboration centers on Advantech adopting the Qualcomm Dragonwing AI on-prem appliance solution, powered by Qualcomm Cloud AI 100 Ultra accelerator into the Advantech SKY-641E3 4U high-performance edge server. By utilizing a specialized PCIe switch backplane, this solution provides the high-bandwidth connectivity required to unlock the full potential of high-density AI inference for enterprise and industrial verticals.
Unmatched Inference Density for Massive Concurrency

Unlike standard dual-slot GPU servers, the SKY-641E3 is architected specifically for maximum inference throughput. By supporting up to 12 single-slot FH/FL accelerators within a single 4U chassis, it offers a superior density-to-footprint ratio. This configuration is optimized for high-concurrency edge workloads, such as city-wide intelligent video analytics or large-scale retail AI agents, allowing for the simultaneous processing of hundreds of real-time data streams without the thermal overhead of power-hungry training GPUs.

Panmnesia, an AI infrastructure link solution provider, today announced the signing of a strategic Memorandum of Understanding (MOU) with SK Telecom, South Korea's largest telco and a leading AI company. The agreement, signed at MWC26 in Barcelona, aims to jointly develop a CXL-based next-generation AI data center architecture.

As large-scale AI services continue to expand, data centers are investing heavily in massive deployments of high-performance GPUs, resulting in astronomical costs. Recognizing the need for sustainable scalability, SK Telecom and Panmnesia are focusing beyond simple GPU expansion to technologies that enable more efficient utilization of existing computing resources. Through this collaboration, the two companies aim to simultaneously improve cost efficiency and performance by innovating data center interconnect architecture based on Compute Express Link (CXL) technology.