Hopper / Data Center GPU

NVIDIA H100 data center GPU

H100 should be sourced as part of a server platform: form factor, QVL, cooling, power, networking and software compatibility must be checked.

Request priceView specification

Specs

Procurement specifications

The table helps qualify the request quickly. For a commercial offer we verify datasheet, OEM SKU, availability and warranty region.

ArchitectureHopper
Memory80 GB HBM3 (SXM reference)
Memory bandwidth3.35 TB/s
Form factorSXM / PCIe / NVL by SKU
MIGup to 7 MIG instances
WorkloadsAI training / inference / HPC
NVIDIA H100 GPU

Procurement profile

Where this model makes sense

Used for training, inference, HPC and enterprise AI clusters where server architecture is already defined.

  • LLM training and inference;
  • HPC, scientific computing and analytics;
  • HGX, MGX and OEM server nodes supporting H100.

RFQ

What to verify before quoting

  • SXM, PCIe or NVL SKU;
  • server QVL and thermal design;
  • power budget, airflow and PSUs;
  • InfiniBand/Ethernet, ConnectX, NVLink and cables.

Critical limitation

NVIDIA H100

H100 requires data-center infrastructure: rack, cooling, power and supported server platform must be confirmed before quoting.

Specification sources

The table helps qualify the request quickly. For a commercial offer we verify datasheet, OEM SKU, availability and warranty region.

NVIDIA H100 GPU

Internal links

Related NVIDIA pages

These links keep model-level pages inside the current catalog architecture without overloading the global menu.

Data Center GPUs

NVIDIA H200 for LLM and AI infrastructure

 Data Center GPUs

NVIDIA B200 / GB200 for AI factories

 Data Center GPUs

NVIDIA DGX H100 / H200

Back to section

RFQ

Request price and availability

Attach SKU, BOM or model list. For server systems, include site, power, cooling, network and compatible server details.