Can I Run Gemma 4 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)?
作者： Jakub Rusinowski · 最后更新： 2026年6月3日
Yes
Yes — Gemma 4 31B at Q6_K needs about 28.1 GB of the 32 GB usable on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM) (~3.9 GB spare), at ~44.8 tok/s (estimated), with room for about 32,768 tokens of context.
Confidence: high · Recommended quantization: Q6_K · Estimated speed: ~44.8 tok/s
See what else this hardware can run →
作为亚马逊联盟成员，我们从符合条件的购买中获得收入。云 GPU 链接为推荐链接——我们可能获得佣金，您无需额外付费。
RTX 5090 Desktop (32 GB VRAM, 64 GB RAM) — what it gives a model
| Usable memory for models | 32 GB |
| Memory bandwidth | 1792 GB/s |
| Form factor | Desktop |
| Operating system | Windows or Linux |
| Memory upgradeable | Yes |
Gemma 4 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM): memory by quantization
| Quant | Memory needed | Fits 32 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 64.7 GB | ✗ No | — | — | 62 GB |
| Q8_0 | 35.7 GB | ✗ No | — | — | 32.9 GB |
| Q6_K | 28.1 GB | ✓ Yes | 32K | ~44.8 tok/s | 25.4 GB |
| Q5_K_M | 24.7 GB | ✓ Yes | 64K | ~50.6 tok/s | 22 GB |
| Q4_K_M | 21.4 GB | ✓ Yes | 128K | ~57.6 tok/s | 18.7 GB |
| Q3_K_M | 15.9 GB | ✓ Yes | 128K | ~75.4 tok/s | 13.2 GB |
| Q2_K | 12.9 GB | ✓ Yes | 128K | ~90.8 tok/s | 10.2 GB |
Which Gemma 4 sizes fit
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| Gemma 4 31B | 21.4 GB | ✓ Fits | ~57.6 tok/s |
| Gemma 4 26B-A4B | 17 GB | ✓ Fits | ~230.4 tok/s |
| Gemma 4 12B (Unified) | 8.7 GB | ✓ Fits | ~124.1 tok/s |
| Gemma 4 E4B | 6.8 GB | ✓ Fits | ~159.6 tok/s |
| Gemma 4 E2B | 4.9 GB | ✓ Fits | ~203.6 tok/s |
RTX 5090 desktop limitations
- 575 W board power — budget for a 1000 W+ PSU and the heat it puts into the room.
- Models larger than 32 GB must offload to system RAM, which costs roughly an order of magnitude in speed.
Recommended setup
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving
How these numbers are calculated
- Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
- 32 GB of VRAM on the NVIDIA GeForce RTX 5090 at 1792 GB/s.
- 64 GB of system RAM available for CPU offload when a model exceeds VRAM.
- Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
- KV cache is computed from this model's published attention configuration.
FAQ
Can I run Gemma 4 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)?
Yes — Gemma 4 31B at Q6_K needs about 28.1 GB of the 32 GB usable on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM) (~3.9 GB spare), at ~44.8 tok/s (estimated), with room for about 32,768 tokens of context.
Which quantization of Gemma 4 should I use on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)?
Q6_K — it needs about 28.1 GB of the 32 GB available, downloads as roughly 25.4 GB, and runs at an estimated 44.8 tokens/sec with up to 32K of context.
What limits Gemma 4 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)?
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
Which runtime should I use?
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving
Other Computers
Other Models on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)
- GLM-4.7 / GLM-Z1 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)
- GLM-5 / GLM-5.1 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)
- GPT-OSS on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)
- Granite 3.0 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)
- IBM Granite 4.0 on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)
Gemma 4 on GPUs
- Gemma 4 on NVIDIA GeForce RTX 5090
- Gemma 4 on NVIDIA GeForce RTX 5080
- Gemma 4 on NVIDIA GeForce RTX 5070 Ti
- Gemma 4 on NVIDIA GeForce RTX 5070