🔌
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 CogVLM-17B?
Yes — here's how fast and up to what context, with real measured numbers.
CogVLM-17B on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 33k | 28.7-34.9 t/s | ~0.8-0.9 (slow) |
| Q8 | yes | 98k | 57.4-69.7 t/s | ~1.5-1.9 (slow) |
| Q4 | yes | 126k | 102.4-124.5 t/s | ~2.7-3.3 (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 FuriosaAI RNGD run CogVLM-17B?▶
Yes. At Q4 it generates ~102.4-124.5 tok/s and fits up to 126k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~2.7-3.3 tok/s.
How much context fits?▶
At Q4, up to about 126k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.