🔌
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 Hermes 3 8B?
Yes — here's how fast and up to what context, with real measured numbers.
Hermes 3 8B on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2899k | 917.4-1121.3 t/s | ~1.6-1.9 (slow) |
| Q8 | yes | 2960k | 1834.9-2242.6 t/s | ~3.2-3.9 (slow) |
| Q4 | yes | 2987k | 3276.6-4004.7 t/s | ~5.6-6.9 (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 Hermes 3 8B?▶
Yes. At Q4 it generates ~3276.6-4004.7 tok/s and fits up to 2987k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~5.6-6.9 tok/s.
How much context fits?▶
At Q4, up to about 2987k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.