Can I Run gemma 4 31B on a AMD Ryzen AI Max 390 (64GB)?
Runs at Q8_0 — near-lossless quality.
10 quantizations fit your 48.0GB
| Quant | Min VRAM | Recommended | File size | Headroom |
|---|---|---|---|---|
| Q8_0BEST | 35.7 GB | 37.2 GB | 32.6 GB | +12.3 GB |
| Q6_K | 27.9 GB | 29.4 GB | 25.2 GB | +20.1 GB |
| Q5_K_M | 24.2 GB | 25.7 GB | 21.9 GB | +23.8 GB |
| Q5_K_S | 23.6 GB | 25.1 GB | 21.3 GB | +24.4 GB |
| Q4_K_M | 20.8 GB | 22.3 GB | 18.7 GB | +27.2 GB |
| Q4_K_S | 19.7 GB | 21.2 GB | 17.8 GB | +28.3 GB |
| Q3_K_L | 15.6 GB | 17.1 GB | 16.6 GB | +32.5 GB |
| Q3_K_M | 14.7 GB | 16.2 GB | 15.3 GB | +33.3 GB |
| Q3_K_S | 13.6 GB | 15.1 GB | 13.8 GB | +34.4 GB |
| Q2_K | 11.8 GB | 13.3 GB | 11.9 GB | +36.3 GB |
Try it in the cloud first
Don't want to download gemma 4 31B just to try it? Use a hosted API or rent a GPU by the second.
Affiliate links — we earn a commission at no cost to you.
All quant variants, benchmark scores, and use-case tags.
Top-ranked open-source models that fit in 48.0GB.
FAQ
Can the AMD Ryzen AI Max 390 (64GB) run gemma 4 31B?
Yes. The AMD Ryzen AI Max 390 (64GB)'s 48.0GB of unified memory is enough to run gemma 4 31B at Q8_0 quantization (35.7GB required).
What's the best quantization to use?
Q8_0 is the highest-precision quantization that fits in your 48.0GB. It uses about 35.7GB of memory and 37.2GB recommended for comfortable inference.
What if I need more headroom for context length?
KV cache memory grows with context length. The numbers above assume a baseline 2K-4K context. For long-context use (32K+), add another 2-6GB depending on the model architecture.