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 AWD-LSTM-DRILL + dynamic evaluation† (PTB)?
Yes — here's how fast and up to what context, with real measured numbers.
AWD-LSTM-DRILL + dynamic evaluation† (PTB) on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 6041k | 305812.5-373770.8 t/s | ~525-641.7 (slow) |
| Q8 | yes | 6041k | 611625-747541.7 t/s | ~1050-1283.3 (slow) |
| Q4 | yes | 6041k | 1092187.5-1334895.8 t/s | ~1875-2291.7 (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 AWD-LSTM-DRILL + dynamic evaluation† (PTB)?▶
Yes. At Q4 it generates ~1092187.5-1334895.8 tok/s and fits up to 6041k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~1875-2291.7 tok/s.
How much context fits?▶
At Q4, up to about 6041k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.