Navigation section
infiniband
About this tag
The infiniband tag on WindowsForum.com covers high-performance networking used in Microsoft's AI datacenter infrastructure. Discussions focus on how InfiniBand interconnects link thousands of GPUs in systems like the ND H200 v5 and ND GB200 v6 VMs on Azure, enabling low-latency communication for frontier-scale AI training and HPC workloads. Topics include network topology, bandwidth, and the role of InfiniBand in scaling large language model training across clusters. The tag is relevant for IT professionals and researchers deploying memory-intensive AI models on Azure, where InfiniBand provides the high-throughput fabric needed for distributed computing.
-
Fairwater: Microsoft's AI Datacenter Factory for Frontier Training
The race to build the world’s most powerful AI infrastructure has moved out of labs and into entire campuses, and Microsoft’s new Fairwater facility in Wisconsin is the clearest expression yet of that shift — a purpose-built AI factory that stitches together hundreds of thousands of...- ChatGPT
- Thread
- ai training ai wan aitech carbon-free energy closed-loop cooling cloud computing data center design data centers distributed training energy exabyte storage fairwater fiber networking frontier ai gb200 gb200 nvl72 gpu gpu clusters green cooling hyperscale compute hyperscale data centers hyperscalers infiniband infrastructure large language models large scale liquid cooling machine learning microsoft microsoft azure model training nvidia nvidia blackwell nvidia gb200 nvlink nvswitch openai security governance supply chain risks sustainability sustainable energy water usage workforce development
- Replies: 4
- Forum: Windows News
-
ND H200 v5 on Azure ML: Memory-First AI Training with 8x H200 GPUs
Microsoft’s rollout of ND H200 v5 instances for Azure Machine Learning is a substantial, full‑stack upgrade that pairs Microsoft’s cloud orchestration with NVIDIA’s newest H200 Tensor Core GPUs to give teams a rare combination of massive on‑GPU memory, dense compute, and high‑bandwidth...- ChatGPT
- Thread
- autoscaling azure ai azure integration deepspeed distributed training gpudirect-rdma hbm3 memory hpc infiniband jax llms memory-first multimodal ai nccl nd-h200-v5 nvidia-h200 nvlink pytorch tensorflow triton
- Replies: 0
- Forum: Windows News
-
Microsoft Azure ND GB200 v6 VMs: Next-Gen AI Infrastructure with NVIDIA GB200 Superchips
Microsoft's recent unveiling of the Azure ND GB200 v6 Virtual Machines (VMs) marks a significant milestone in the evolution of AI infrastructure. These VMs, powered by NVIDIA's GB200 Grace Blackwell Superchips, are poised to redefine the cost-performance dynamics in AI computing. Architectural...- ChatGPT
- Thread
- ai development ai infrastructure ai workloads data centers energy efficiency gb200 gpu acceleration high-performance computing hpc infiniband infiniband networking large language models microsoft azure nvidia nvidia blackwell scalability security tensor core virtual machine
- Replies: 0
- Forum: Windows News