🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN ON DATACENTER GPU
Run ProBERTa on Datacenter GPU hardware
13 options ranked by estimated speed at Q4 / 8k context.
| Hardware | Est tok/s | Fits 8k? | Max ctx | ~Price |
|---|---|---|---|---|
| NVIDIA B200 → | ~282,467.5 | yes | 2672k | — |
| AMD Instinct MI325X → | ~187,500 | yes | 3570k | — |
| NVIDIA H200 → | ~169,480.5 | yes | 1956k | — |
| AMD Instinct MI300X 192GB → | ~165,625 | yes | 2672k | ~$15,000 |
| NVIDIA H100 SXM5 80GB → | ~118,283.3 | yes | 1100k | ~$30,000 |
| Intel Gaudi 3 → | ~108,116.9 | yes | 1774k | — |
| NVIDIA A100 SXM4 80GB → | ~71,993.9 | yes | 1100k | ~$15,000 |
| NVIDIA H100 PCIe 80GB → | ~70,616.9 | yes | 1100k | ~$25,000 |
| NVIDIA A100 PCIe 80GB → | ~68,321.8 | yes | 1100k | ~$12,000 |
| NVIDIA RTX PRO 6000 Blackwell → | ~63,272.7 | yes | 1324k | ~$8,500 |
| NVIDIA RTX 6000 Ada 48GB → | ~33,896.1 | yes | 651k | ~$6,800 |
| NVIDIA L40S 48GB → | ~30,506.5 | yes | 651k | ~$10,000 |
| AMD Radeon Pro W7900 48GB → | ~27,000 | yes | 651k | ~$3,999 |
Estimate at Q4 (memory-bound). Click any option for its full per-hardware estimate and measured numbers.
FAQ
What Datacenter GPU hardware runs ProBERTa?▶
On GenAIList, 13 Datacenter GPU options can run ProBERTa at Q4 — the fastest is the NVIDIA B200 at ~282,467.5 tok/s.
How fast is ProBERTa on Datacenter GPU hardware?▶
Estimated generation speed ranges from ~27,000 to ~282,467.5 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.