Post Syndicated from John Lee original https://www.servethehome.com/nextsilicon-maverick-2-brings-dataflow-and-hbm3e-to-hpc-customers/
The NextSilicon Maverick-2 brings dataflow architecture with HBM3E memory to HPC customers, including a Sandia National Labs win
The post NextSilicon Maverick-2 Brings Dataflow and HBM3e to HPC Customers appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/amd-and-openai-ink-megadeal-for-6gw-of-future-ai-compute/
AMD and OpenAI inked a megadeal for AI compute covering 6GW of compute including 1GW of MI450 targeting 2H 2026 deployment
The post AMD and OpenAI Ink Megadeal for 6GW of Future AI Compute appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/nvidia-rubin-cpx-is-an-ai-gpu-for-next-gen-nvidia-ai-gpus/
The NVIDIA Rubin CPX is an AI GPU for next-gen NVIDIA GPUs. Planned for 2026, NVIDIA will have heterogeneous GPU and memory types clustered
The post NVIDIA Rubin CPX is an AI GPU for Next-Gen NVIDIA AI GPUs appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/the-new-d-matrix-jetstream-400g-ethernet-card-for-data-center-scale-ai-inference/
The new d-Matrix Jetstream 400G card is designed to help the company scale out its Corsair AI inference platfrom using lower-cost switching
The post The New d-Matrix JetStream 400G Ethernet Card for Data Center Scale AI Inference appeared first on ServeTheHome.
Post Syndicated from Ryan Smith original https://www.servethehome.com/googles-ironwood-tpu-swings-for-reasoning-model-leadership-at-hot-chips-2025/
Closing out the machine learning sessions at Hot Chips 2025 is Google, who is at the show to talk about their latest tensor processing unit (TPU), codenamed Ironwood. Revealed by the company a few months ago, Ironwood is the first Google TPU that is explicitly designed for large-scale AI inference (rather than AI training). Paired […]
The post Google’s Ironwood TPU Swings for Reasoning Model Leadership at Hot Chips 2025 appeared first on ServeTheHome.
Post Syndicated from Ryan Smith original https://www.servethehome.com/amd-dives-deep-on-cdna-4-architecture-and-mi350-accelerator-at-hot-chips-2025/
The second big machine learning accelerator talk of the afternoon belongs to AMD. The company’s chip architects are at this year’s show to tell the audience all about the CDNA 4 architecture, which is powering AMD’s new MI350 family of accelerators. Like it’s MI300 predecessor, AMD is using 3D die stacking to build up a […]
The post AMD Dives Deep on CDNA 4 Architecture and MI350 Accelerator at Hot Chips 2025 appeared first on ServeTheHome.
Post Syndicated from Ryan Smith original https://www.servethehome.com/huawei-presents-ub-mesh-interconnect-for-large-ai-supernodes-at-hot-chips-2025/
The third and final machine learning presentation before the afternoon break comes from Huawei. Unlike many of the other ML vendors who are here to pitch products, Huawei’s presentation is more focused on fundamental technology. In this case, how to use efficiently use meshes to interconnect the chips within large AI systems. Eyeing so-called SuperNodes […]
The post Huawei Presents UB-Mesh Interconnect for Large AI SuperNodes at Hot Chips 2025 appeared first on ServeTheHome.
Post Syndicated from Ryan Smith original https://www.servethehome.com/d-matrix-presents-corsair-an-in-memory-computing-architecture-for-inference-at-hot-chips-2025/
The second machine learning presentation of the afternoon comes from d-Matrix. The company specializes in hardware for AI inference, and as of late has been tackling the matter of how to improve inference performance by using in-memory computing. Along those lines, the company is presenting their Corsair in-memory computing chiplet architecture at Hot Chips. Not […]
The post d-Matrix Presents Corsair, An In-Memory Computing Architecture For Inference, at Hot Chips 2025 appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/rebellions-rebel-quad-ucie-and-144tb-hbm3e-accelerator-at-hot-chips-2025/
At Hot Chips 2025, we saw a live demo of the Rebellions REBEL-Quad, an AI accelerator with four ASICs, 144GB of HBM3E, and more using UCIe
The post Rebellions REBEL-Quad UCIe and 144TB HBM3E Accelerator at Hot Chips 2025 appeared first on ServeTheHome.
Post Syndicated from Cliff Robinson original https://www.servethehome.com/new-faster-amd-alveo-v80-accelerator-with-hbm2e-and-fast-networking/
The new, faster, AMD Alveo V80 packs 32GB of HBM2e memory, four QSFP56 200G ports, and many other features with the SoC on a pre-built card
The post New Faster AMD Alveo V80 Accelerator with HBM2e and Fast Networking appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/the-2025-pcie-gpu-in-server-guide-supermicro-nvidia/
In our 2025 PCIe GPU in Server Guide we get into different types of PCIe GPUs and what types of servers they are used in
The post The 2025 PCIe GPU in Server Guide appeared first on ServeTheHome.
Post Syndicated from Eric Smith original https://www.servethehome.com/this-is-the-amd-instinct-mi350/
We saw the AMD Instinct MI350 package at AMD Advancing AI 2025 this week in its package form and in the UBB 8-GPU platform
The post This is the AMD Instinct MI350 appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/not-just-for-oreos-and-tailers-amd-helios-next-gen-ai-racks-go-double-wide/
The AMD Helios with the MI450 will scale to 31TB of memory and 1.4PB/s of memory bandwidth per rack. It is so big, it needs bigger racks
The post Not Just for Oreos and Tailers AMD Helios Next-Gen AI Racks Go Double-Wide appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/amd-instinct-mi350-launch-event-coverage/
The new AMD Instinct MI350 series scales up to 1.4kW and 288GB of HBM3E per GPU with a focus on AI performance
The post AMD Instinct MI350 Launch Event Coverage appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/amd-mi350-and-cdna-4-architecture-launched-with-rocm-7/
AMD provided some new details around its CDNA 4 architecture, the AMD MI350 construction, and launched ROCm 7
The post AMD MI350 and CDNA 4 Architecture Launched with ROCm 7 appeared first on ServeTheHome.
Post Syndicated from Cliff Robinson original https://www.servethehome.com/micron-begins-shipping-hbm4-memory-for-next-gen-ai/
Micron HBM4 is now shipping to customers, offering a big bump in performance. These are designed for 2026 generation AI accelerators
The post Micron Begins Shipping HBM4 Memory for Next-Gen AI appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/gigabyte-g383-r80-aap1-amd-instinct-mi300a-server-review/
We check out the GIGABYTE G383-R80-AAP1 an AMD Instinct MI300A supercomputer server with a unique APU architecture
The post GIGABYTE G383-R80-AAP1 AMD Instinct MI300A Server Review appeared first on ServeTheHome.
Post Syndicated from Will Taillac original https://www.servethehome.com/qualcomm-discrete-qualcomm-npu-spotted-at-in-dell-pro-max-plus-laptop-at-dtw/
At Dell Tech World 2025, the company showed off a new laptop class that had Qualcomm NPUs instead of a dGPU for heavier AI workloads
The post Qualcomm Discrete NPU Spotted at in Dell Pro Max Plus laptop at DTW appeared first on ServeTheHome.
Post Syndicated from John Lee original https://www.servethehome.com/maxsun-intel-arc-pro-b60-dual-gpu-48gb-at-computex-2025/
We spotted the Maxsun Intel Arc Pro B60 dual GPU 48GB card at Computex 2025 and can show you how this AI focused GPU works
The post Maxsun Intel Arc Pro B60 Dual GPU 48GB at Computex 2025 appeared first on ServeTheHome.
Post Syndicated from Patrick Kennedy original https://www.servethehome.com/intel-arc-pro-b50-and-b60-for-lower-cost-pro-gpus-and-18a-pather-lake-shown-at-computex-2025/
The Intel Arc Pro B50 is a 70W SFF GPU at $299 with 16GB of memory and the B60 is a step up with support for AI inference and virtualization
The post Intel Arc Pro B50 and B60 For Lower Cost Pro GPUs and 18A Pather Lake Shown at Computex 2025 appeared first on ServeTheHome.