New: connect Claude & other AIs to GenAIList over MCP โ research the catalog and contribute to the shared knowledge base. Learn how โ
Run Falcon-OCR 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 โ | ~41,428.6 | yes | 2670k | โ |
| AMD Instinct MI325X โ | ~27,500 | yes | 3568k | โ |
| NVIDIA H200 โ | ~24,857.1 | yes | 1954k | โ |
| AMD Instinct MI300X 192GB โ | ~24,291.7 | yes | 2670k | ~$15,000 |
| NVIDIA H100 SXM5 80GB โ | ~17,348.2 | yes | 1098k | ~$30,000 |
| Intel Gaudi 3 โ | ~15,857.1 | yes | 1771k | โ |
| NVIDIA A100 SXM4 80GB โ | ~10,559.1 | yes | 1098k | ~$15,000 |
| NVIDIA H100 PCIe 80GB โ | ~10,357.1 | yes | 1098k | ~$25,000 |
| NVIDIA A100 PCIe 80GB โ | ~10,020.5 | yes | 1098k | ~$12,000 |
| NVIDIA RTX PRO 6000 Blackwell โ | ~9,280 | yes | 1322k | ~$8,500 |
| NVIDIA RTX 6000 Ada 48GB โ | ~4,971.4 | yes | 648k | ~$6,800 |
| NVIDIA L40S 48GB โ | ~4,474.3 | yes | 648k | ~$10,000 |
| AMD Radeon Pro W7900 48GB โ | ~3,960 | yes | 648k | ~$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 Falcon-OCR?โถ
On GenAIList, 13 Datacenter GPU options can run Falcon-OCR at Q4 โ the fastest is the NVIDIA B200 at ~41,428.6 tok/s.
How fast is Falcon-OCR on Datacenter GPU hardware?โถ
Estimated generation speed ranges from ~3,960 to ~41,428.6 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.