Dmitry Safronov’s Post

Introducing the GIGABYTE G593-ZX1 and ZX2 🌐🚀 Unleash the Power of #AI Training and Inference with cutting-edge features: ✔️ Supports AMD Instinct™ MI300X architecture, meticulously designed for seamless #AI training and inference ✔️ Top-tier performance and reliability, the foundation for your high-performance computing needs ✔️ Elevate your computing experience to new heights with the paralleled processing power of dual AMD EPYC ✔️ Massive RAM Capacity for your most demanding #AI workloads ✔️ High-Speed Connectivity, ensuring lightning-fast data transfer and efficient server management Are you ready to elevate your #AI and #HPC experience? Order here 👇 https://lnkd.in/eHHy5tgG #AI #HPC #GIGABYTE #Innovation #TechAdvancements #Happyware #HappywareServer #ServerSolutions #Tech #AITraining #ArtificialIntelligence #HighPerformanceComputing #TechnologyTrends #AMD #GPGPU #AIInference

Exciting news in the world of storage technology! Our partner PAC Storage has just released the PS5024 U.2 NVMe hybrid flash storage series, a significant advancement for AI and HPC applications. With a 5th Gen Xeon processor, PCIe 5.0, and NVMe-oF integration, the PS5024UE can double throughput performance compared to its predecessor, achieving 50GB/s throughput and 1.3 million IOPS with a latency of 0.3ms, facilitated by 100GbE connectivity. This high-performance, low-latency storage solution is ideal for extensive dataset processing workloads with simultaneous random and sequential data access requirements. If you're looking for contemporary AI and HPC needs, the PS5024UE stands out as a game-changer in the storage industry. #PACStorage #StorageTechnology #AI #HPC #DataProcessing #WCT #datastorage #mediaandentertainment #government #gov #datacenter

Clustered Systems: Explore the new supercomputing infrastructure at #SC24 ✨ On-site at #SC24, John Furrier digs into clustered systems as one of the main themes of the supercomputing conference, emphasizing their capabilities to support computational workloads that #GenAI models demand. “Clustered systems aggregate multiple computing units such as servers, GPUs, and networking components into a single, unified computing environment. This allows the systems to handle the kind of computational workloads that generative AI models demand, whether it’s training massive language models or running inference across diverse applications,” Furrier writes. He furthers,saying “Nvidia has been at the forefront of this shift, not just with their GPUs but with their entire ecosystem—NVLink, CUDA, and their DGX systems. It’s not just Nvidia driving this forward. Companies like Dell Technologies are advancing the infrastructure landscape with their PowerEdge XE9680 servers, which integrate Intel Gaudi 3 AI accelerators, and next-generation PowerEdge systems featuring AMD EPYC 5th Gen processors.” 📰 Read the complete article, and join Furrier for live SC24 analysis at theCUBE.net https://lnkd.in/dNw7ufj7 #EnterpriseAI #HPC

I know have recently been banging on about the current importance of thinking about modernising your #virtualised and #containerised environments. Opex reduction, alternative virtualisation, application moderisation and mobility - true hybrid cloud.... YOUR hybrid cloud. Joe 's post highlights the rubber hitting the road, not only data services and storage persistence on prem, but also core application GRUNT! AI inferencing or creating bespoke business personalised outcomes safely using your secret sauce data in your place of control not someone or something else's is the domain of IBM Fusion. True busines #AI and the race to be relevant in this space is heating up... IBM is the leader in promoting and creating platforms and solutions that deliver AI trustworthiness for your business use. Lets have a chat. #Fusion #RHOS Red Hat #teampurple #watsonx #ibmstorage Andrew Gee Barry Whyte Andrew Beattie Sheldon Dargaville Grant Smith Reginald D'Souza Frank Lee Dave McDonnell ☁️ Ben Harrison ☁️ Justin Harrington

At SDC 24, MangoBoost and AMD showed how to supercharge AI storage—because speed is everything in the world of GPUs. MangoBoost’s Chief Product Officer, Eriko Nurvitadhi, and AMD’s Storage and Networking Architect, Craig Carlson, showcased how the MangoBoost DPU fully accelerates direct Ethernet-based communication between AMD GPUs and remote, disaggregated storage servers using NVMe-over-TCP and peer-to-peer technology. The presentation covered the latest AI trends, such as the increasing demand for high-speed storage driven by large language models (LLMs) and massive datasets. They also highlighted how DPUs can optimize GPU system efficiency in accessing storage, and shared case studies demonstrating significant performance gains and reduced CPU utilization in AI workloads using AMD’s MI300X GPUs and open-source ROCm software. For those eager to see this groundbreaking solution in real life, be on the lookout for the upcoming OCP Global Summit in San Jose and the Supercomputing Conference in Atlanta! #DataProcessingUnit #DPU #Storage #NVMe #TCP #P2P #AMD #GPU #MI300X #ROCm #DataCenter

The power of #GPUs can unlock advanced computing use cases, but how you deploy them matters cc Jason Lee & Umesh Shah

MangoBoost and AMD co-presented a technical seminar at SNIA Developer Conference (SDC) in Santa Clara last week. "Accelerating GPU Server Access to Network-Attached Disaggregated Storage using Data Processing Unit (DPU)" The title of the talk says it all. The talk covered several real-world case studies of accelerating top-of-the-line GPU and storage systems using MangoBoost's GPU Storage Boost technology. https://lnkd.in/gK39rVAF

IBM Storage Fusion HCI System has released a new GPU server today that takes a huge step forward in the support of AI applications. The new server, called the G03, can be configured to have as many as 8 double-wide GPU adapter cards.  Fine-tuning and inferencing with foundation models used in AI applications increasingly requires faster and more powerful hardware, and the Fusion HCI G03 meets that need head on! #ibmstorage #ibmstoragefusion #storagefusionhci #gpu #openshift https://lnkd.in/gs5pdza7

Supermicro yesterday launched our latest JBOF supporting either 36 E3.S SSDs or 24 U.2 SSDs, with a raw capacity of up to 1.44 PB, later supporting up to 2 PB with newer 60TB SSDs. The dense design is ideal for both scale-up and scale-out file and object storage, while the high networking throughput up to 800Gb/s per JBOF—supports the needs of AI training and HPC workloads.

The “An Introduction to the Compute Express Link (CXL) Interconnect” white paper, written by Debendra Das Sharma and Rob Blankenship (Intel Corporation) and Daniel S. Berger (Microsoft and University of Washington), explores the growth of CXL technology, covering the CXL 1.1, 2.0, and 3.X specifications, as well as the technology’s current and future impacts on the data center landscape. Read the white paper to learn more about the important use cases and applications enabled by CXL technology, including highly efficient accelerators, server memory bandwidth and capacity expansion, memory pooling and sharing, and efficient peer-to-peer communication: https://bit.ly/3WCthEx #CXLConsortium #ComputeExpressLink #CXL #Datacenter