New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Can the Intel Gaudi 3 run MoEFold2D?
Yes — here's how fast and up to what context, with real measured numbers.
MoEFold2D on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1774k | 1117708.3-1387500 t/s | ~12083.3-15000 (slow) |
| Q8 | yes | 1774k | 2235416.7-2775000 t/s | ~24166.7-30000 (slow) |
| Q4 | yes | 1774k | 3991815.5-4955357.1 t/s | ~43154.8-53571.4 (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 MoEFold2D?▶
Yes. At Q4 it generates ~3991815.5-4955357.1 tok/s and fits up to 1774k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~43154.8-53571.4 tok/s.
How much context fits?▶
At Q4, up to about 1774k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.