io.net’s Post

📯New Release!🎊 Introducing the 技嘉 R163-S30, R163-S32, and E163-S30 series, a total of six server models fully upgraded! 💡🚀 🤘Supporting Gen 4th/5th 英特爾公司 Xeon Scalable Processors, this series boasts stable and powerful performance, offering a variety of memory configurations to meet corporate needs, and now includes titanium PSU, providing a strong and stable energy backup for HPC.⚡️🎯 🦾The R163/E163 series is particularly suited for AI and HPC applications, with Intel's specially designed embedded HBM (High Bandwidth Memory) offering unprecedented power boosts for high-performance computing. Whether it's data centers, cloud computing, or other workloads requiring powerful processing capabilities, the R163/E163 will deliver outstanding solutions.😍 Upgrade now 🌟 https://gct.pse.is/56fphd #GIGABYTE #GIGABYTEgroup #GigaComputing #GIGABYTEServer #serversolutions #RackServer #TitaniumPSU #AI #HPC

Interesting stat on how networking is commandeering a bigger portion of the cloud & infra spend in the AI computing environments. > The pricing for 200 Gb/sec and 400 Gb/sec InfiniBand fabrics proves it, where it is not uncommon for the network to comprise 20 percent or more of the cost of a cluster in HPC and hyperscale/cloud markets where customers are used to spending under 10 percent on the interconnect. HPC and AI vendors like Cray, NVIDIA, AWS, Microsoft, AMD, and others knew this all along. This has driven their acquisition and investment strategies over the past few years. Question remains -- how quickly will another standard like UEC establish itself as a viable alternative for the AI Computing space? I know the design-in for UEC and others will be dependent on not only availability of products on the market, but also the maturity and adoption of alternatives to IB verbs, OFED & PSM like OFA's libfabric. Source: https://lnkd.in/gsemgh4d cc: Balaji Anand Jim

In the competitive realm of artificial intelligence, dominated by Nvidia’s formidable GPUs, a bold newcomer is betting on a different contender. Nscale, a little-known GPU cloud provider based in Norway, has launched a new AI Cloud Services platform built around AMD’s MI300X accelerators. The move is a bold gambit to challenge Nvidia’s dominance and offer customers a potentially cheaper and greener alternative for running AI workloads in the cloud. - #datacenter - https://lnkd.in/e6C3bVYZ

💼 Big Moves in the Data Center Industry! 💼 AMD is set to acquire hyperscale server maker ZT Systems for $4.9 billion, following a previous $1 billion investment. The acquisition, expected to close in the first half of 2025, will see AMD focusing on ZT’s systems design capabilities, while spinning off its manufacturing assets. This move strengthens AMD's AI strategy, aiming to deliver cutting-edge AI infrastructure for cloud and enterprise customers. ZT Systems, known for its work with AWS and Microsoft Azure, brings valuable expertise in systems design and rack-scale solutions to AMD. With this acquisition, AMD is positioning itself to compete more fiercely in the AI space, combining its Epyc CPUs and Instinct GPUs with ZT’s design prowess to deliver comprehensive data center AI solutions. #DataCenters #AI #AMD #ZTSystems #CloudComputing

AMD, the chipmaker hot on the heels of Nvidia in the AI race, today announced a big acquisition to boost its position as an “ecosystem” partner for companies building big AI businesses: It is acquiring ZT Systems, which provides compute design and infrastructure for AI, cloud and general purpose computing, for $4.9 billion. The deal is a mix of cash and stock and includes a contingent payment of up to $400 million if certain performance metrics are met. The plan is to incorporate ZT Systems’ computing infrastructure design business. AMD said it will look to sell ZT Systems’ data center infrastructure manufacturing business to “a strategic partner. https://lnkd.in/gi7cjhcj ZT Systems AMD Blog AI and Cloud Storage – Pure Storage - https://lnkd.in/g_FeRmWy

CUDOS has announced their exciting collaboration with Storj and VALDI a division of Storj. This brings multi-petabyte capacity on-chain compute, leveraging Storj’s advanced S3-compatible storage solutions within the CUDOS network. Key highlights: 🌟CUDOS is an exclusive cloud partner of NVIDIA (NCP), delivering market-leading H100 and H200 chips. 🌟Storj’s distributed storage platform enhances security through client-side encryption and data sharding. 🌟Files are encrypted, split into fragments, and distributed globally across thousands of storage nodes in over 100 countries. 🌟Enhanced security, privacy, durability, availability, and superior edge performance. "Partnering with Storj allows us to offer an unparalleled blend of compute and storage capabilities," said Matt Hawkins, CEO at CUDOS. "This collaboration enhances our ability to deliver high-performance solutions to leading AI applications and demanding enterprise organizations." Ben Golub, CEO at Storj, added, "Our distributed storage solutions ideally complement CUDOS's compute infrastructure. Together, we’re setting a new standard for scalable, secure, and sustainable services." Discover how our partnership can revolutionize your AI and Web3 projects. https://hubs.li/Q02LVVTf0 #AIcompute #Web3 #NVIDIA #DePIN

Distributed compute distributed storage is the winning combo for AI modelling and inference work. And..market/customer momentum is proving this out. Very excited by our partnership with CUDOS

Getting your AI Infrastructure right the first time is a breeze with Liqid!  #Liqid  #ComposableInfrastructure  #Partnership  #slurm #kubernetes  #data #delltechnologies #dellemc #nvidiavgpu #nvidia #machinelearning #ml #Liqid #GPU #AI #HPC

This announcement is big! Trillium TPUs achieve an impressive 4.7X increase in peak compute performance per chip compared to TPU v5e. We doubled the High Bandwidth Memory (HBM) capacity and bandwidth, and also doubled the Interchip Interconnect (ICI) bandwidth over TPU v5e. Additionally, Trillium is equipped with third-generation SparseCore, a specialized accelerator for processing ultra-large embeddings common in advanced ranking and recommendation workloads. Trillium TPUs make it possible to train the next wave of foundation models faster and serve those models with reduced latency and lower cost. Critically, our sixth-generation TPUs are also our most sustainable: Trillium TPUs are over 67% more energy-efficient than TPU v5e.

Amazon CloudWatch Container Insights with Enhanced Observability for EKS now auto-discovers critical health metrics from your AWS accelerators Trainium and Inferentia, and AWS high performance network adapters (Elastic Fabric Adapters) as well as NVIDIA GPUs. You can visualize these out-of-the-box metrics in curated Container Insights dashboards to help monitor your accelerated infrastructure and optimize your AI workloads for operational excellence.