🔌
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 OmniGen?
Yes — here's how fast and up to what context, with real measured numbers.
OmniGen on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 5926k | 1931.4-2360.7 t/s | ~3.3-4.1 (slow) |
| Q8 | yes | 5984k | 3862.9-4721.3 t/s | ~6.6-8.1 (slow) |
| Q4 | yes | 6009k | 6898-8430.9 t/s | ~11.8-14.5 (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 OmniGen?▶
Yes. At Q4 it generates ~6898-8430.9 tok/s and fits up to 6009k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~11.8-14.5 tok/s.
How much context fits?▶
At Q4, up to about 6009k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.