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 ProBERTa?
Yes — here's how fast and up to what context, with real measured numbers.
ProBERTa on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1773k | 24386.4-30272.7 t/s | ~263.6-327.3 (slow) |
| Q8 | yes | 1773k | 48772.7-60545.5 t/s | ~527.3-654.5 (slow) |
| Q4 | yes | 1774k | 87094.2-108116.9 t/s | ~941.6-1168.8 (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 ProBERTa?▶
Yes. At Q4 it generates ~87094.2-108116.9 tok/s and fits up to 1774k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~941.6-1168.8 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.