🔌
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 InternVL2_5-78B?
Yes — here's how fast and up to what context, with real measured numbers.
InternVL2_5-78B on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 730k | 93.6-114.4 t/s | ~0.2-0.2 (slow) |
| Q8 | yes | 969k | 187.2-228.8 t/s | ~0.3-0.4 (slow) |
| Q4 | yes | 1074k | 334.3-408.6 t/s | ~0.6-0.7 (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 InternVL2_5-78B?▶
Yes. At Q4 it generates ~334.3-408.6 tok/s and fits up to 1074k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~0.6-0.7 tok/s.
How much context fits?▶
At Q4, up to about 1074k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.