NVIDIA H100 80GB
Hopper-generation datacenter GPU. aicenterAI does not invent tokens/s figures; compare providers on listed $/hr and confirm throughput on your workload.
VRAM
80 GB
TDP
700 W
Architecture
Hopper
No throughput (tokens/s) is listed — we do not hard-code unsourced performance claims.
Cloud offers
- RunPod · $2.69/hr · global · Rent on RunPod
- DigitalOcean · $3.39/hr · nyc2 · Rent on DigitalOcean