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 GLA Transformer 340M?
Yes — here's how fast and up to what context, with real measured numbers.
GLA Transformer 340M on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1764k | 3155.9-3917.6 t/s | ~34.1-42.4 (slow) |
| Q8 | yes | 1769k | 6311.8-7835.3 t/s | ~68.2-84.7 (slow) |
| Q4 | yes | 1771k | 11271-13991.6 t/s | ~121.8-151.3 (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 GLA Transformer 340M?▶
Yes. At Q4 it generates ~11271-13991.6 tok/s and fits up to 1771k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~121.8-151.3 tok/s.
How much context fits?▶
At Q4, up to about 1771k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.