What GPU do you need to run DeepSeek R1 0528?
671B params 160K max context Reasoning
The smallest GPU memory that runs DeepSeek R1 0528 well is 512 GB (Q4_K_M, e.g. M3 Ultra). 2 of 132 GPUs run it well.
MMLU 90.8 SWE-bench Verified 57.6
VRAM needed per quantization
| Quantization | Weights | 16K context | 32K context | Full 160K |
|---|---|---|---|---|
| Q4_K_M | 409.2 GB | 410.3 GB | 411.3 GB | 419.6 GB |
| Q8_0 | 713.3 GB | 714.3 GB | 715.4 GB | 723.7 GB |
Weights plus fp16 KV cache. Leave ~1 GB headroom for the inference engine.
GPUs that can run DeepSeek R1 0528
Best quantization that fits on each GPU, with the context and speed it gets. For chips with several memory sizes, the smallest that works is shown.