🔌
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 Intel Arc Pro B50 run DeciLM 6B?
Yes — here's how fast and up to what context, with real measured numbers.
DeciLM 6B on the Intel Arc Pro B50
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 14k | 11.4-14.1 t/s | ~2-2.5 (slow) |
| Q8 | yes | 57k | 22.8-28.3 t/s | ~4.1-5.1 (slow) |
| Q4 | yes | 77k | 40.7-50.5 t/s | ~7.3-9 (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 Intel Arc Pro B50 run DeciLM 6B?▶
Yes. At Q4 it generates ~40.7-50.5 tok/s and fits up to 77k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~7.3-9 tok/s.
How much context fits?▶
At Q4, up to about 77k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.