🔌
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 Gaudi 3 run Refact-1.6B?
Yes — here's how fast and up to what context, with real measured numbers.
Refact-1.6B on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1725k | 670.6-832.5 t/s | ~7.3-9 (slow) |
| Q8 | yes | 1750k | 1341.3-1665 t/s | ~14.5-18 (slow) |
| Q4 | yes | 1760k | 2395.1-2973.2 t/s | ~25.9-32.1 (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 Gaudi 3 run Refact-1.6B?▶
Yes. At Q4 it generates ~2395.1-2973.2 tok/s and fits up to 1760k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~25.9-32.1 tok/s.
How much context fits?▶
At Q4, up to about 1760k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.