New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Can the FuriosaAI RNGD run BART-IT?
Yes — here's how fast and up to what context, with real measured numbers.
BART-IT on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 647k | 3482.1-4232.1 t/s | ~92.9-112.9 (slow) |
| Q8 | yes | 649k | 6964.3-8464.3 t/s | ~185.7-225.7 (slow) |
| Q4 | yes | 650k | 12436.2-15114.8 t/s | ~331.6-403.1 (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 BART-IT?▶
Yes. At Q4 it generates ~12436.2-15114.8 tok/s and fits up to 650k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~331.6-403.1 tok/s.
How much context fits?▶
At Q4, up to about 650k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.