🔌
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 Apple M3 Max 36GB run GemNet-T (OC20)?
Yes — here's how fast and up to what context, with real measured numbers.
GemNet-T (OC20) on the Apple M3 Max 36GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 471k | 53684.2-64736.8 t/s | ~1789.5-2157.9 (slow) |
| Q8 | yes | 471k | 107368.4-129473.7 t/s | ~3578.9-4315.8 (slow) |
| Q4 | yes | 471k | 191729.3-231203 t/s | ~6391-7706.8 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the Apple M3 Max 36GB run GemNet-T (OC20)?▶
Yes. At Q4 it generates ~191729.3-231203 tok/s and fits up to 471k context; beyond that it spills to SSD swap and slows to ~6391-7706.8 tok/s.
How much context fits?▶
At Q4, up to about 471k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.