M3 Ultra 512GB vs NVIDIA RTX 5070 Ti for local LLMs
41 reports on the M3 Ultra 512GB and 53 on the NVIDIA RTX 5070 Ti
Which is faster for local LLMs?
On Gemma 4 26B · 4B active at 4-bit, the M3 Ultra 512GB has 87.8 generation tokens/s (one run); prompt processing 2595.3 tokens/s (one run); the NVIDIA RTX 5070 Ti has 137.7 generation tokens/s (one run); prompt processing 5951.0 tokens/s (one run). Those are community reports whose engines and settings differ, so the gap is a rough sign, not a controlled measure of the two cards. Memory bandwidth, which caps how fast a card can write, is 9% higher on the NVIDIA RTX 5070 Ti (819 GB/s on the M3 Ultra 512GB, 896 on the NVIDIA RTX 5070 Ti).
- Memory
- 512GB unified
- Memory bandwidth
- 819 GB/s
- FP16 compute
- 28.4 TFLOPS
- Reports
- 41
- Memory
- 16GB
- Memory bandwidth
- 896 GB/s
- FP16 compute
- 176 TFLOPS
- Reports
- 53
Measured on both
Generation and prompt processing speeds from plain runs (one device, one request, no speculative decoding, the whole model in the device's memory) of the same model size at the same bit class. Each side shows the engines and reported context behind its median. Context can mean a configured window or prompt depth, so these are community comparisons with differing settings. How to read these
| Model | M3 Ultra 512GB | NVIDIA RTX 5070 Ti |
|---|---|---|
| Gemma 4 26B · 4B active4-bit | 87.8one runPP 2595.3 (one run)llama.cppcontext not stated | 137.7one runPP 5951.0 (one run)llama.cppcontext not stated |
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. Where a model was also measured above, the measurement wins.
| Model | M3 Ultra 512GB | NVIDIA RTX 5070 Ti |
|---|---|---|
| Gemma 4 26B · 4B active | 21.5 t/sFP16 | 130.3 t/sQ3_K_S |
| Qwen3.8 27B | 8.0 t/sFP16 | 58.4 t/sQ2_K |
| Qwen3.8 125B · 6B active | 13.6 t/sFP16 | doesn't fit at these settings |
| Qwen3.6 35B · 3B active | 28.7 t/sFP16 | 193.2 t/sQ2_K |
| DeepSeek V4 Flash 284B · 13B active | 22.6 t/sQ8_0 | doesn't fit at these settings |
| Qwen3.6 27B | 8.0 t/sFP16 | 58.4 t/sQ2_K |
| DeepSeek V4.1 Flash 552B · 16B active | 30.0 t/sQ4_K_M | doesn't fit at these settings |
Change the context, RAM or model in the calculator: M3 Ultra 512GB · NVIDIA RTX 5070 Ti
Every report on each card, with its source: M3 Ultra 512GB · NVIDIA RTX 5070 Ti