M5 Ultra 256GB vs M5 Ultra 512GB for local LLMs
30 reports on the M5 Ultra 256GB and 1 on the M5 Ultra 512GB
Which is faster for local LLMs?
Nobody has reported the same model at the same quant on both cards yet, so there's no measured head-to-head. Memory bandwidth, which caps how fast a card can write, is about the same on both (1,229 GB/s on the M5 Ultra 256GB, 1,229 on the M5 Ultra 512GB). The estimates below compare them on the models people run most.
M5 Ultra 256GB
- Memory
- 256GB unified
- Memory bandwidth
- 1,229 GB/s
- FP16 compute
- not on record
- Reports
- 30
M5 Ultra 512GB
- Memory
- 512GB unified
- Memory bandwidth
- 1,229 GB/s
- FP16 compute
- not on record
- Reports
- 1
Estimated by the calculator
The quant the calculator recommends for each card and its estimated speed at a 32,768-token context with 32 GB of system RAM, from memory bandwidth and the model's shape.
| Model | M5 Ultra 256GB | M5 Ultra 512GB |
|---|---|---|
| Qwen3.8 27B | 14.9 t/sFP16 | 14.9 t/sFP16 |
| Qwen3.8 125B · 6B active | 50.8 t/sQ8_0 | 25.4 t/sFP16 |
| Qwen3.6 35B · 3B active | 53.5 t/sFP16 | 53.5 t/sFP16 |
| DeepSeek V4 Flash 284B · 13B active | 66.9 t/sQ4_K_M | 42.1 t/sQ8_0 |
| Qwen3.6 27B | 14.9 t/sFP16 | 14.9 t/sFP16 |
| Gemma 4 26B · 4B active | 40.1 t/sFP16 | 40.1 t/sFP16 |
Change the context, RAM or model in the calculator: M5 Ultra 256GB · M5 Ultra 512GB
Every report on each card, with its source: M5 Ultra 256GB · M5 Ultra 512GB