NVIDIA H100 SXM5 80GB HBM2e GPU Accelerator for HPE

NVIDIA H100 SXM5 80GB | Hopper Architecture Data Centre GPU

The world's most powerful AI training GPU — built on NVIDIA Hopper architecture with SXM5 form factor for maximum NVLink bandwidth. Designed for large-scale AI/ML training, HPC, and generative AI. Available in Qatar for enterprise and government deployments.

Key Specifications

• GPU Architecture: NVIDIA Hopper (GH100)
• GPU Memory: 80GB HBM2e
• Memory Bandwidth: 3.35 TB/s
• FP8 Tensor Core: 1,979 TFLOPS
• FP16 Tensor Core: 989 TFLOPS
• FP64 (Tensor Core): 67 TFLOPS
• TDP: 700W
• Form Factor: SXM5 (board-mounted)
• NVLink: 900 GB/s bidirectional
• Multi-Instance GPU (MIG): Up to 7 instances

Ideal Workloads

• Large Language Model (LLM) Training — GPT, LLaMA, Falcon
• Generative AI & Diffusion Models
• High-Performance Computing (HPC)
• Scientific Simulation & Research

Compatible Systems

• HPE ProLiant DL380 Gen10 Plus (HGX H100 configuration)
• HPE Cray EX235a Accelerator Blade

Send your RFQ to sales@serverparts.qa or use the RFQ form on our website.