🔌
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 ProtGPT2?
Yes — here's how fast and up to what context, with real measured numbers.
ProtGPT2 on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 6019k | 9945.1-12155.1 t/s | ~17.1-20.9 (slow) |
| Q8 | yes | 6030k | 19890.2-24310.3 t/s | ~34.1-41.7 (slow) |
| Q4 | yes | 6035k | 35518.3-43411.2 t/s | ~61-74.5 (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 ProtGPT2?▶
Yes. At Q4 it generates ~35518.3-43411.2 tok/s and fits up to 6035k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~61-74.5 tok/s.
How much context fits?▶
At Q4, up to about 6035k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.