{"slug": "instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu-gpu-collaboration", "title": "Instinctive and EPYC; VAST Data sets up widespread AMD CPU/GPU collaboration", "summary": "VAST Data is expanding its collaboration with AMD to support AMD EPYC and Instinct processors in its AI OS software, aiming to build and operate high-performance AI factories at scale. The deal includes VAST selecting 6th-gen AMD EPYC processors for its CBox and EBox platforms, and developing an AI Infrastructure Reference Architecture with AMD and DriveNets. VAST Data VP John Mao said the industry is discovering that inference is fundamentally a data problem, while AMD Corporate VP Derek Dicker stated the collaboration enables customers to accelerate inference and deploy AI at scale without compromise.", "body_md": "# Instinctive and EPYC; VAST Data sets up widespread AMD CPU/GPU collaboration\n\n[VAST Data](https://www.blocksandfiles.com/ai-ml/2026/07/15/vast-data-and-cloudera-offer-combined-ai-factory/5271629) is setting up a deal with AMD for its AI OS software to support AMD EPYC and Instinct processors, and reference hardware to use EPYC processors as well, and aiming to build and operate high-performance, AMD-based, AI factories at scale.\n\nVAST Data has a tight and comprehensive relationship with the AI industry’s main GPU supplier, Nvidia, and Nvidia Neocloud and enterprise customers. Now it is extending its AI ecosystem coverage to better include AMD, Nvidia’s main GPU competitor, apart from in-house hyperscaler accelerator chips. VAST and AMD say that they are expanding their collaboration to deliver an open and flexible approach to AI infrastructure that combines accelerated computing, intelligent data services and optimized inference software into a unified platform for AI clouds and enterprise AI deployments.\n\nJohn Mao, VP, Global Technology Alliances at VAST Data, said: “AI is entering an operational phase where infrastructure efficiency matters as much as model performance. The industry is discovering that inference is fundamentally a data problem. Success depends on how effectively organizations can bring data, compute, memory and intelligence together as a single system. The VAST AI Operating System was built for this transition, giving AI cloud providers and enterprises a more efficient, scalable and open foundation for training, inference and the next generation of agentic AI applications.”\n\nEchoing this, Derek Dicker, Corporate VP, Enterprise Business Group at AMD, stated: “Our expanded collaboration with VAST combines AMD EPYC CPUs and Instinct GPUs with the software foundation customers need to accelerate inference, improve infrastructure efficiency and deploy AI at scale. Together, we’re enabling AI clouds and enterprises to build high-performance AI factories without compromise.”\n\nThe enhanced collaboration features:\n\nVAST selecting 6\n\nthGen AMD EPYC processors (formerly codenamed Venice) to power the 6th-generation of CBox and 3rd-generation of EBox platforms to underpin the VAST[AI OS](https://www.blocksandfiles.com/ai-ml/2026/02/26/vast-broadens-ai-platform-push-with-nvidia-tie-up-and-control-plane/4092639). Gen 6 EPYC supports PCIe Gen-6 that, compared to 50 percent slower PCIe Gen 5, enables 2X the I/O bandwidth and lowers latency for AI data services including database, data warehouse, and event streaming via VAST’s DataBase and DataEngine capabilities.An AI Infrastructure Reference Architecture developed by VAST, AMD and DriveNets including AMD Helios rack-scale AI infrastructure, the VAST AI OS, and DriveNets' AI Fabric networking. These reference architectures document infrastructure support for model training, inference, reinforcement learning (RL) and KV cache workloads with sizing considerations and guidance for AI cloud providers and enterprises working on AI factory deployments.\n\nExpanded ecosystem collaboration with suppliers including TensorMesh and EmbeddedLLM to accelerate deployment of production-ready inference architectures optimized for agentic AI applications.\n\nNew KV cache and inference optimizations that combine AMD Instinct GPUs, AMD Infinity Context, and AMD ROCm software with the VAST AI OS.\n\nAutomated KV Cache Lifecycle Management: Purging cached data that contains sensitive or personal information is a critical enterprise compliance challenge. The integration leverages VAST’s native data lifecycle policies to automatically expire and delete KV cache data, helping ensure robust security, privacy, and regulatory compliance capabilities without manual operational overhead.\n\nThe AMD\n\n[Pensando](https://www.blocksandfiles.com/nvme/2022/09/07/pensando-gets-dpu-flying-with-vsphere-8/1605178)Pollara 400 AI NIC provides the high-performance data path connecting AMD Instinct GPUs to the VAST AI OS. Using NFS over TCP and NFS over RDMA, this integration efficiently moves data from GPU memory to the NVMe SSD-based VAST storage cluster, enabling the KV-cache and storage access that large-scale inference and agentic AI workloads depend on.Proven deployments across leading AI cloud providers delivering AMD technology-powered AI services to customers around the world, including 5C, Core42, Crusoe, EmbeddedLLM, Phanos.AI, TensorWave, and Vultr.\n\nEarly VAST testing utilizing an AMD Instinct MI355X GPU demonstrated 9X speedup in time-to-first-token (TTFT), and 9.7X more token throughput utilizing VAST for KV Cache offloading with high concurrency agentic AI workloads. The testing evaluated KV cache offloading to local host RAM versus offloading to a remote VAST Data partition over NFS/RDMA.\n\nBecause speedups are always relative to the underlying hardware, running these workloads on a GPU baseline with lower compute capacity may yield an even higher relative speedup compared to the latency of reading directly from storage.\n\nVAST will showcase the joint architecture, reference designs and inference optimization technologies at AMD Advancing AI 2026 in San Francisco on July 22-23.", "url": "https://wpnews.pro/news/instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu-gpu-collaboration", "canonical_source": "https://www.blocksandfiles.com/flash/2026/07/23/instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu/gpu-collaboration/5276173", "published_at": "2026-07-23 17:30:00+00:00", "updated_at": "2026-07-23 17:36:20.866132+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-infrastructure", "ai-chips", "ai-products", "ai-startups"], "entities": ["VAST Data", "AMD", "John Mao", "Derek Dicker", "DriveNets", "TensorMesh", "EmbeddedLLM", "Pensando"], "alternates": {"html": "https://wpnews.pro/news/instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu-gpu-collaboration", "markdown": "https://wpnews.pro/news/instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu-gpu-collaboration.md", "text": "https://wpnews.pro/news/instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu-gpu-collaboration.txt", "jsonld": "https://wpnews.pro/news/instinctive-and-epyc-vast-data-sets-up-widespread-amd-cpu-gpu-collaboration.jsonld"}}