DeepSeek on AMD Strix Halo
Can I run DeepSeek-R1-Distill-Qwen 32B on an Ryzen AI Max+ 395 128GB?
Yes. DeepSeek-R1-Distill-Qwen 32B at Q4_K_M uses 21.2 GB of the 96.0 GB available on a Ryzen AI Max+ 395 128GB, and runs at about 6.84 tokens per second. You can push the context to 128K.
Runs comfortably
21.2 GB of 96.0 GB · 22%096 GB
Weights
18.6 GB
KV cache
2.0 GB
Runtime overhead
0.5 GB
DeepSeek-R1-Distill-Qwen 32B at Q4_K_M leaves 74.8 GB spare on a Ryzen AI Max+ 395 128GB. There is room to raise the context length or move up a quantisation level.
Generation6.84tok/s
Prompt processing321tok/s
Max context128Ktokens
KV per 1K tokens0GB
Every quantisation of DeepSeek-R1-Distill-Qwen 32B on a Ryzen AI Max+ 395 128GB
Highlighted row is the highest quality that still fits at 8K context.
| Quantisation | bpw | Weights | Total | Verdict | Max ctx | tok/s |
|---|---|---|---|---|---|---|
| FP16 / BF16 | 16.00 | 61.0 GB | 63.6 GB | Runs comfortably | 128K | 2.17 |
| INT8 / W8A8 | 8.50 | 32.1 GB | 34.6 GB | Runs comfortably | 128K | 4.06 |
| Q8_0 (GGUF) | 8.50 | 32.1 GB | 34.6 GB | Runs comfortably | 128K | 4.06 |
| FP8 (E4M3) | 8.00 | 30.5 GB | 33.0 GB | Runs comfortably | 128K | 4.26 |
| Q6_K | 6.56 | 25.0 GB | 27.6 GB | Runs comfortably | 128K | 5.16 |
| Q5_K_M | 5.67 | 21.6 GB | 24.2 GB | Runs comfortably | 128K | 5.93 |
| Q5_K_S | 5.52 | 21.1 GB | 23.6 GB | Runs comfortably | 128K | 6.08 |
| Q4_K_M | 4.85 | 18.6 GB | 21.2 GB | Runs comfortably | 128K | 6.84 |
| AWQ 4-bit | 4.25 | 18.3 GB | 20.9 GB | Runs comfortably | 128K | 6.93 |
| GPTQ 4-bit | 4.25 | 18.3 GB | 20.9 GB | Runs comfortably | 128K | 6.93 |
| MXFP4 | 4.25 | 18.3 GB | 20.9 GB | Runs comfortably | 128K | 6.93 |
| Q4_K_S | 4.58 | 17.6 GB | 20.2 GB | Runs comfortably | 128K | 7.20 |
| Q4_0 | 4.55 | 17.5 GB | 20.1 GB | Runs comfortably | 128K | 7.24 |
| IQ4_XS | 4.25 | 16.4 GB | 19.0 GB | Runs comfortably | 128K | 7.69 |
| Q3_K_M | 3.91 | 15.2 GB | 17.7 GB | Runs comfortably | 128K | 8.27 |
| IQ3_M | 3.70 | 14.4 GB | 17.0 GB | Runs comfortably | 128K | 8.68 |
| IQ3_XXS | 3.06 | 12.1 GB | 14.7 GB | Runs comfortably | 128K | 10.2 |
| Q2_K | 2.63 | 10.6 GB | 13.1 GB | Runs comfortably | 128K | 11.6 |
| IQ2_XXS | 2.06 | 8.5 GB | 11.0 GB | Runs comfortably | 128K | 14.1 |
| IQ1_M | 1.75 | 7.4 GB | 9.9 GB | Runs comfortably | 128K | 16.0 |