🔌
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 Instinct MI455X run Evo 2 40B?
Yes — here's how fast and up to what context, with real measured numbers.
Evo 2 40B on the Instinct MI455X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 962k | 153.2-187.2 t/s | ~0.3-0.4 (slow) |
| Q8 | yes | 1085k | 306.4-374.5 t/s | ~0.6-0.8 (slow) |
| Q4 | yes | 1139k | 547.1-668.7 t/s | ~1.1-1.4 (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 Instinct MI455X run Evo 2 40B?▶
Yes. At Q4 it generates ~547.1-668.7 tok/s and fits up to 1139k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~1.1-1.4 tok/s.
How much context fits?▶
At Q4, up to about 1139k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.