🔌
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 NVIDIA GeForce RTX 3080 10GB run Unified-IO (XL)?
Yes — here's how fast and up to what context, with real measured numbers.
Unified-IO (XL) on the NVIDIA GeForce RTX 3080 10GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 28k | 94.8-113 t/s | ~5-5.9 (slow) |
| Q8 | yes | 73k | 189.7-226.1 t/s | ~10-11.9 (slow) |
| Q4 | yes | 92k | 338.7-403.7 t/s | ~17.8-21.2 (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 NVIDIA GeForce RTX 3080 10GB run Unified-IO (XL)?▶
Yes. At Q4 it generates ~338.7-403.7 tok/s and fits up to 92k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~17.8-21.2 tok/s.
How much context fits?▶
At Q4, up to about 92k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.