NVIDIA HGX H100- 8GPU(8x NVIDIA H100 SXM) India

The NVIDIA HGX™ H100 8-GPU baseboard is the gold standard for AI supercomputing clusters. It links eight H100 SXM GPUs using NVSwitch technology, allowing every GPU to talk to every other GPU at full NVLink speed (900 GB/s).

With a total of 640GB HBM3 memory, it is architected for massively parallel workloads like training Large Language Models (LLMs) and mixture-of-experts (MoE) models. This platform serves as the fundamental node for the world’s largest AI clouds and supercomputers, delivering the bisection bandwidth required to scale workloads across thousands of GPUs.