🔌
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 Intel Arc A770 16GB run FragLlama: Next-fragment prediction for molecular design?
Yes — here's how fast and up to what context, with real measured numbers.
FragLlama: Next-fragment prediction for molecular design on the Intel Arc A770 16GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | no | — | — | — |
| Q8 | yes | 47k | 46.4-57.6 t/s | ~3.3-4.1 (slow) |
| Q4 | yes | 71k | 82.9-102.9 t/s | ~5.9-7.3 (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.
→ All models on the Intel Arc A770 16GB
→ FragLlama: Next-fragment prediction for molecular design specs
→ Other hardware
FAQ
Can the Intel Arc A770 16GB run FragLlama: Next-fragment prediction for molecular design?▶
Yes. At Q4 it generates ~82.9-102.9 tok/s and fits up to 71k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~5.9-7.3 tok/s.
How much context fits?▶
At Q4, up to about 71k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.