Data & Infrastructure

H100 / H200

NVIDIA data-center GPUs based on Hopper architecture. H100 (2022) and H200 (2024) are the workhorses of frontier-model training and inference through 2026. Predecessors to Blackwell (B100/B200) which shipped in volume 2024-2025.

Related terms

TensorRT-LLM
NVIDIA's inference stack for LLMs, optimized for NVIDIA GPUs. Includes quantization, speculative decoding, and multi-GPU tensor parallelism. Fastest option for open-weights inference on NVIDIA hardware; less flexible on other accelerators.

Next steps

Take the assessment— See where you stand

Where this fits

H100 / H200 is part of the Data & Infrastructure vocabulary used in the Generative AI Maturity Framework. See the full glossary for the complete set of 149 defined terms, or take the free maturity assessment to see where your organisation stands.

Related terms

Related on this site

Framework dimensions

Free tools

Whitepapers

Next steps

Where this fits