🔌
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 Gemma 3 1B?
Yes — here's how fast and up to what context, with real measured numbers.
Gemma 3 1B on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 6011k | 7339.5-8970.5 t/s | ~12.6-15.4 (slow) |
| Q8 | yes | 6026k | 14679-17941 t/s | ~25.2-30.8 (slow) |
| Q4 | yes | 6033k | 26212.5-32037.5 t/s | ~45-55 (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 Gemma 3 1B?▶
Yes. At Q4 it generates ~26212.5-32037.5 tok/s and fits up to 6033k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~45-55 tok/s.
How much context fits?▶
At Q4, up to about 6033k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.