🔌
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 MI430X run xLAM-8x22B?
Yes — here's how fast and up to what context, with real measured numbers.
xLAM-8x22B on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 348k | 52.1-63.6 t/s | ~0.1-0.1 (slow) |
| Q8 | yes | 778k | 104.1-127.2 t/s | ~0.2-0.2 (slow) |
| Q4 | yes | 967k | 185.9-227.2 t/s | ~0.3-0.4 (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 MI430X run xLAM-8x22B?▶
Yes. At Q4 it generates ~185.9-227.2 tok/s and fits up to 967k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~0.3-0.4 tok/s.
How much context fits?▶
At Q4, up to about 967k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.