The NVIDIA H800 SXM5 packs 80GB of HBM3 with 3360 GB/s of memory bandwidth, making it one of the fastest cards you can run a local model on. Here is which models fit, how fast they run, and what quantisation gives the best balance of quality and speed.
Already have it? Find the best model for it. Still choosing? See what a budget buys. Or the full specifications.
The NVIDIA H800 SXM5 has 80GB VRAM and 3360 GB/s memory bandwidth, making it one of the fastest cards you can run a local model on. This guide covers which LLMs fit, expected tok/s performance, and recommended settings.
See the full NVIDIA H800 SXM5 specs page for detailed specifications, all compatible models, and speed estimates.