New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Can the NVIDIA A100 SXM4 80GB run Pangu-Weather?
Yes — here's how fast and up to what context, with real measured numbers.
Pangu-Weather on the NVIDIA A100 SXM4 80GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1092k | 2907.2-3464.7 t/s | ~57-68 (slow) |
| Q8 | yes | 1096k | 5814.3-6929.4 t/s | ~114.1-135.9 (slow) |
| Q4 | yes | 1098k | 10382.7-12374 t/s | ~203.7-242.7 (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 A100 SXM4 80GB run Pangu-Weather?▶
Yes. At Q4 it generates ~10382.7-12374 tok/s and fits up to 1098k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~203.7-242.7 tok/s.
How much context fits?▶
At Q4, up to about 1098k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.