🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the NVIDIA B200 run CARP?
Yes — here's how fast and up to what context, with real measured numbers.
CARP on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2653k | 4541.2-5412.1 t/s | ~22.7-27.1 (slow) |
| Q8 | yes | 2663k | 9082.4-10824.3 t/s | ~45.4-54.1 (slow) |
| Q4 | yes | 2667k | 16218.6-19329 t/s | ~81.1-96.6 (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 CARP?▶
Yes. At Q4 it generates ~16218.6-19329 tok/s and fits up to 2667k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~81.1-96.6 tok/s.
How much context fits?▶
At Q4, up to about 2667k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.