🔌
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 M2 Max 96GB run data2vec (language)?
Yes — here's how fast and up to what context, with real measured numbers.
data2vec (language) on the Apple M2 Max 96GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1274k | 192.9-232.6 t/s | ~4.8-5.8 (slow) |
| Q8 | yes | 1285k | 385.7-465.2 t/s | ~9.6-11.6 (slow) |
| Q4 | yes | 1289k | 688.8-830.6 t/s | ~17.2-20.8 (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 M2 Max 96GB run data2vec (language)?▶
Yes. At Q4 it generates ~688.8-830.6 tok/s and fits up to 1289k context; beyond that it spills to SSD swap and slows to ~17.2-20.8 tok/s.
How much context fits?▶
At Q4, up to about 1289k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.