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 Rita-XLarge?
Yes — here's how fast and up to what context, with real measured numbers.
Rita-XLarge on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2636k | 2433.3-2900 t/s | ~12.2-14.5 (slow) |
| Q8 | yes | 2654k | 4866.7-5800 t/s | ~24.3-29 (slow) |
| Q4 | yes | 2662k | 8690.5-10357.1 t/s | ~43.5-51.8 (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 Rita-XLarge?▶
Yes. At Q4 it generates ~8690.5-10357.1 tok/s and fits up to 2662k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~43.5-51.8 tok/s.
How much context fits?▶
At Q4, up to about 2662k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.