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 TrellisNet-MoS (1.4x larger) PTB?
Yes — here's how fast and up to what context, with real measured numbers.
TrellisNet-MoS (1.4x larger) PTB on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2671k | 85882.4-102352.9 t/s | ~429.4-511.8 (slow) |
| Q8 | yes | 2672k | 171764.7-204705.9 t/s | ~858.8-1023.5 (slow) |
| Q4 | yes | 2672k | 306722.7-365546.2 t/s | ~1533.6-1827.7 (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 TrellisNet-MoS (1.4x larger) PTB?▶
Yes. At Q4 it generates ~306722.7-365546.2 tok/s and fits up to 2672k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~1533.6-1827.7 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.