🔌
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 Pythia-6.9b?
Yes — here's how fast and up to what context, with real measured numbers.
Pythia-6.9b on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1231k | 423.2-504.3 t/s | ~2.1-2.5 (slow) |
| Q8 | yes | 1284k | 846.4-1008.7 t/s | ~4.2-5 (slow) |
| Q4 | yes | 1307k | 1511.4-1801.2 t/s | ~7.6-9 (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 Pythia-6.9b?▶
Yes. At Q4 it generates ~1511.4-1801.2 tok/s and fits up to 1307k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~7.6-9 tok/s.
How much context fits?▶
At Q4, up to about 1307k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.