🔌
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 M4 Max 36GB run TAPE Transformer?
Yes — here's how fast and up to what context, with real measured numbers.
TAPE Transformer on the Apple M4 Max 36GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 470k | 4885.3-5891.1 t/s | ~89.5-107.9 (slow) |
| Q8 | yes | 471k | 9770.5-11782.1 t/s | ~178.9-215.8 (slow) |
| Q4 | yes | 471k | 17447.4-21039.5 t/s | ~319.5-385.3 (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 M4 Max 36GB run TAPE Transformer?▶
Yes. At Q4 it generates ~17447.4-21039.5 tok/s and fits up to 471k context; beyond that it spills to SSD swap and slows to ~319.5-385.3 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.