🔌
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 FuriosaAI RNGD run Qwen-VLA?
Yes — here's how fast and up to what context, with real measured numbers.
Qwen-VLA on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 247k | 94.7-115 t/s | ~2.5-3.1 (slow) |
| Q8 | yes | 286k | 189.3-230.1 t/s | ~5-6.1 (slow) |
| Q4 | yes | 303k | 338.1-410.9 t/s | ~9-11 (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 FuriosaAI RNGD run Qwen-VLA?▶
Yes. At Q4 it generates ~338.1-410.9 tok/s and fits up to 303k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~9-11 tok/s.
How much context fits?▶
At Q4, up to about 303k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.