New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Can the NVIDIA B200 run UniRep?
Yes — here's how fast and up to what context, with real measured numbers.
UniRep on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2672k | 160439.6-191208.8 t/s | ~802.2-956 (slow) |
| Q8 | yes | 2672k | 320879.1-382417.6 t/s | ~1604.4-1912.1 (slow) |
| Q4 | yes | 2672k | 572998.4-682888.5 t/s | ~2865-3414.4 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to system RAM offload (depends on your system RAM) and speed drops to the "beyond" figure.
FAQ
Can the NVIDIA B200 run UniRep?▶
Yes. At Q4 it generates ~572998.4-682888.5 tok/s and fits up to 2672k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~2865-3414.4 tok/s.
How much context fits?▶
At Q4, up to about 2672k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.