Rent a GPU

Can the NVIDIA RTX 3090 run Gemma 3 27B?

RTX 3090 × Gemma 3 27B8K context
YesRuns fully on the GPU
Memory at Q4_K_M21.4 GB of 24 GB
Weights 16.6 GBKV cache 4.2 GBRuntime 0.6 GB
QuantQ4_K_MGood, most popular
Memory21.4 GBof 24 GB usable
Speed~34tokens per second
  1. FP16
  2. Q8_0
  3. Q6_K
  4. Q5_K_M
  5. Q4_K_M
  6. Q3_K_M
  7. Q2_K

Yes. The NVIDIA RTX 3090 can run Gemma 3 27B fully in VRAM at Q4_K_M (good, most popular), using about 21.4 GB of the 24 GB available, at roughly 34 tokens/second.

Try other context lengths

Run it with Ollama

With Ollama installed, one command downloads and starts Gemma 3 27B:

ollama run gemma3:27b

Ollama's default download is Q4_K_M (about 21.4 GB here). That is also the best quant that fits.

Run it in the cloud

No NVIDIA RTX 3090? Run Gemma 3 27B in the cloud.

Rent a cloud GPU by the hour. No upfront cost, stop any time.

Rent a GPU on RunPodRent a GPU on Vast.ai

Memory needed by quantization

At 8,192 tokens of context. The black line marks the 24 GB the NVIDIA RTX 3090 can use.

QuantQualityMemoryFits?Speed
FP16Full quality59.6 GBToo big—
Q8_0Near-lossless33.9 GBToo big—
Q6_KExcellent27.2 GBToo big—
Q5_K_MVery good24.2 GBToo big—
Q4_K_MGood, most popular21.4 GBFits~34 tok/s
Q3_K_MNoticeably degraded18.2 GBFits~42 tok/s
Q2_KHeavily degraded15 GBFits~55 tok/s

GPUs that run Gemma 3 27B comfortably

Models that run well on the NVIDIA RTX 3090

Frequently asked questions

Can NVIDIA RTX 3090 run Gemma 3 27B?

Yes. The NVIDIA RTX 3090 can run Gemma 3 27B fully in VRAM at Q4_K_M (good, most popular), using about 21.4 GB of the 24 GB available, at roughly 34 tokens/second.

How much VRAM does Gemma 3 27B need?

About 21.4 GB at Q4_K_M with 8,192 tokens of context, and about 59.6 GB at full FP16 precision.

How fast is Gemma 3 27B on the NVIDIA RTX 3090?

Roughly 34 tokens per second at Q4_K_M, estimated from the NVIDIA RTX 3090's 936 GB/s memory bandwidth.