🔌
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 Xinyuan (心元大模型)?
Yes — here's how fast and up to what context, with real measured numbers.
Xinyuan (心元大模型) on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 539k | 76.6-95.1 t/s | ~0.8-1 (slow) |
| Q8 | yes | 624k | 153.3-190.3 t/s | ~1.7-2.1 (slow) |
| Q4 | yes | 662k | 273.7-339.8 t/s | ~3-3.7 (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 Xinyuan (心元大模型)?▶
Yes. At Q4 it generates ~273.7-339.8 tok/s and fits up to 662k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~3-3.7 tok/s.
How much context fits?▶
At Q4, up to about 662k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.