🔌
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 ADM?
Yes — here's how fast and up to what context, with real measured numbers.
ADM on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1757k | 1919.5-2382.8 t/s | ~20.8-25.8 (slow) |
| Q8 | yes | 1765k | 3839-4765.7 t/s | ~41.5-51.5 (slow) |
| Q4 | yes | 1769k | 6855.4-8510.1 t/s | ~74.1-92 (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 ADM?▶
Yes. At Q4 it generates ~6855.4-8510.1 tok/s and fits up to 1769k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~74.1-92 tok/s.
How much context fits?▶
At Q4, up to about 1769k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.