🔌
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 CogVLM-17B?
Yes — here's how fast and up to what context, with real measured numbers.
CogVLM-17B on the Instinct MI455X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1381k | 363.2-443.9 t/s | ~0.7-0.9 (slow) |
| Q8 | yes | 1446k | 726.4-887.8 t/s | ~1.5-1.8 (slow) |
| Q4 | yes | 1474k | 1297.1-1585.3 t/s | ~2.6-3.2 (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 CogVLM-17B?▶
Yes. At Q4 it generates ~1297.1-1585.3 tok/s and fits up to 1474k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~2.6-3.2 tok/s.
How much context fits?▶
At Q4, up to about 1474k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.