🔌
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 LLaVA-NeXT-34B (LLaVA-1.6)?
Yes — here's how fast and up to what context, with real measured numbers.
LLaVA-NeXT-34B (LLaVA-1.6) on the Intel Gaudi 3
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 178k | 30.9-38.3 t/s | ~0.3-0.4 (slow) |
| Q8 | yes | 311k | 61.8-76.7 t/s | ~0.7-0.8 (slow) |
| Q4 | yes | 369k | 110.3-136.9 t/s | ~1.2-1.5 (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 LLaVA-NeXT-34B (LLaVA-1.6)?▶
Yes. At Q4 it generates ~110.3-136.9 tok/s and fits up to 369k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~1.2-1.5 tok/s.
How much context fits?▶
At Q4, up to about 369k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.