£16,183.00 Ex. VAT
Available with a Lead Time: Submit an enquiry for accurate lead time information.
The NVIDIA H100 80 GB, built on the Hopper architecture, is designed for enterprise data-centre environments running large-model AI training, inference workloads, HPC and multi-tenant GPU acceleration.
H100 gives infrastructure teams a high-performance, secure GPU platform for the most demanding AI deployments.
Key technical highlights:
• Architecture: NVIDIA Hopper – 4th-generation Tensor Cores, Transformer Engine support
• Memory: 80 GB HBM2e with ECC
• Bandwidth: up to 2 TB/s
• Compute: FP64 up to 26 TFLOPS, FP32 up to 51 TFLOPS
• Multi-Instance GPU (MIG): supports up to 7 independent partitions
• Form factor & power: dual-slot PCIe (Gen5 x16), typical power up to 350 W
Ideal for: large model training, inference clusters, confidential computing, virtualised AI workloads
This HPE-validated SKU (R9S41C) integrates into enterprise server infrastructure with trusted reliability.
Contact Steel City Consulting today. We’ll evaluate your IT infrastructure’s compatibility with the H100 and provide expert deployment advice tailored to your environment.
