🔌
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 phi-3-mini 3.8B 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 → | ~3,270.7 | yes | 2640k | — |
| AMD Instinct MI325X → | ~2,171.1 | yes | 3538k | — |
| NVIDIA H200 → | ~1,962.4 | yes | 1924k | — |
| AMD Instinct MI300X 192GB → | ~1,917.8 | yes | 2640k | ~$15,000 |
| NVIDIA H100 SXM5 80GB → | ~1,369.6 | yes | 1068k | ~$30,000 |
| Intel Gaudi 3 → | ~1,251.9 | yes | 1742k | — |
| NVIDIA A100 SXM4 80GB → | ~833.6 | yes | 1068k | ~$15,000 |
| NVIDIA H100 PCIe 80GB → | ~817.7 | yes | 1068k | ~$25,000 |
| NVIDIA A100 PCIe 80GB → | ~791.1 | yes | 1068k | ~$12,000 |
| NVIDIA RTX PRO 6000 Blackwell → | ~732.6 | yes | 1292k | ~$8,500 |
| NVIDIA RTX 6000 Ada 48GB → | ~392.5 | yes | 618k | ~$6,800 |
| NVIDIA L40S 48GB → | ~353.2 | yes | 618k | ~$10,000 |
| AMD Radeon Pro W7900 48GB → | ~312.6 | yes | 618k | ~$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 phi-3-mini 3.8B?▶
On GenAIList, 13 Datacenter GPU options can run phi-3-mini 3.8B at Q4 — the fastest is the NVIDIA B200 at ~3,270.7 tok/s.
How fast is phi-3-mini 3.8B on Datacenter GPU hardware?▶
Estimated generation speed ranges from ~312.6 to ~3,270.7 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.