🔌
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 GPT-J-6B?
Yes — here's how fast and up to what context, with real measured numbers.
GPT-J-6B on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 233k | 80.5-97.9 t/s | ~2.1-2.6 (slow) |
| Q8 | yes | 279k | 161.1-195.8 t/s | ~4.3-5.2 (slow) |
| Q4 | yes | 300k | 287.6-349.6 t/s | ~7.7-9.3 (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 GPT-J-6B?▶
Yes. At Q4 it generates ~287.6-349.6 tok/s and fits up to 300k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~7.7-9.3 tok/s.
How much context fits?▶
At Q4, up to about 300k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.