🔌
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 ruDalle: Kandinsky 3.0?
Yes — here's how fast and up to what context, with real measured numbers.
ruDalle: Kandinsky 3.0 on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 564k | 90.2-111.9 t/s | ~1-1.2 (slow) |
| Q8 | yes | 637k | 180.3-223.9 t/s | ~1.9-2.4 (slow) |
| Q4 | yes | 669k | 322-399.8 t/s | ~3.5-4.3 (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 ruDalle: Kandinsky 3.0?▶
Yes. At Q4 it generates ~322-399.8 tok/s and fits up to 669k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~3.5-4.3 tok/s.
How much context fits?▶
At Q4, up to about 669k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.