New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Can the Instinct MI430X run Rita-XLarge?
Yes — here's how fast and up to what context, with real measured numbers.
Rita-XLarge on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 6005k | 6116.3-7475.4 t/s | ~10.5-12.8 (slow) |
| Q8 | yes | 6023k | 12232.5-14950.8 t/s | ~21-25.7 (slow) |
| Q4 | yes | 6031k | 21843.8-26697.9 t/s | ~37.5-45.8 (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 Rita-XLarge?▶
Yes. At Q4 it generates ~21843.8-26697.9 tok/s and fits up to 6031k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~37.5-45.8 tok/s.
How much context fits?▶
At Q4, up to about 6031k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.