🔌
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 MiniCPM-1.2B?
Yes — here's how fast and up to what context, with real measured numbers.
MiniCPM-1.2B on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 613k | 390.8-475 t/s | ~10.4-12.7 (slow) |
| Q8 | yes | 632k | 781.6-949.9 t/s | ~20.8-25.3 (slow) |
| Q4 | yes | 640k | 1395.7-1696.3 t/s | ~37.2-45.2 (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 MiniCPM-1.2B?▶
Yes. At Q4 it generates ~1395.7-1696.3 tok/s and fits up to 640k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~37.2-45.2 tok/s.
How much context fits?▶
At Q4, up to about 640k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.