🔌
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 GLM-10B?
Yes — here's how fast and up to what context, with real measured numbers.
GLM-10B on the Instinct MI430X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2295k | 734-897.1 t/s | ~1.3-1.5 (slow) |
| Q8 | yes | 2356k | 1467.9-1794.1 t/s | ~2.5-3.1 (slow) |
| Q4 | yes | 2382k | 2621.3-3203.8 t/s | ~4.5-5.5 (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 GLM-10B?▶
Yes. At Q4 it generates ~2621.3-3203.8 tok/s and fits up to 2382k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~4.5-5.5 tok/s.
How much context fits?▶
At Q4, up to about 2382k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.