Gemma on NVIDIA Ampere
Can I run Gemma 3 12B on an RTX 3080 10GB?
Yes. Gemma 3 12B at Q4_K_M uses 8.6 GB of the 8.6 GB available on a RTX 3080 10GB, and runs at about 62.1 tokens per second. You can push the context to 8K.
Fits, but tight
8.6 GB of 8.6 GB · 100%09 GB
Weights
7.0 GB
KV cache
0.8 GB
Runtime overhead
0.8 GB
This fits with almost nothing to spare. A background application claiming VRAM will push it over. Drop to the next quantisation down, or quantise the KV cache to Q8_0 — that halves the cache for no meaningful quality loss.
Generation62.1tok/s
Prompt processing1033tok/s
Max context8Ktokens
KV per 1K tokens0GB
Every quantisation of Gemma 3 12B on a RTX 3080 10GB
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 | 22.7 GB | 24.4 GB | Won't fit | — | 2.34 |
| INT8 / W8A8 | 8.50 | 11.8 GB | 13.5 GB | Won't fit | — | 6.90 |
| Q8_0 (GGUF) | 8.50 | 11.8 GB | 13.5 GB | Won't fit | — | 6.90 |
| FP8 (E4M3) | 8.00 | 11.4 GB | 13.0 GB | Won't fit | — | 7.49 |
| Q6_K | 6.56 | 9.3 GB | 10.9 GB | Won't fit | — | 12.1 |
| Q5_K_M | 5.67 | 8.1 GB | 9.7 GB | Won't fit | — | 21.2 |
| Q5_K_S | 5.52 | 7.8 GB | 9.5 GB | Won't fit | — | 23.8 |
| AWQ 4-bit | 4.25 | 7.4 GB | 9.0 GB | Won't fit | 896 | 35.1 |
| GPTQ 4-bit | 4.25 | 7.4 GB | 9.0 GB | Won't fit | 896 | 35.1 |
| MXFP4 | 4.25 | 7.4 GB | 9.0 GB | Won't fit | 896 | 35.1 |
| Q4_K_M | 4.85 | 7.0 GB | 8.6 GB | Fits, but tight | 8K | 62.1 |
| Q4_K_S | 4.58 | 6.6 GB | 8.2 GB | Fits, but tight | 14K | 65.1 |
| Q4_0 | 4.55 | 6.6 GB | 8.2 GB | Fits, but tight | 14K | 65.4 |
| IQ4_XS | 4.25 | 6.2 GB | 7.8 GB | Fits, but tight | 21K | 69.1 |
| Q3_K_M | 3.91 | 5.7 GB | 7.4 GB | Runs comfortably | 28K | 73.9 |
| IQ3_M | 3.70 | 5.5 GB | 7.1 GB | Runs comfortably | 32K | 77.2 |
| IQ3_XXS | 3.06 | 4.6 GB | 6.3 GB | Runs comfortably | 45K | 89.2 |
| Q2_K | 2.63 | 4.1 GB | 5.7 GB | Runs comfortably | 54K | 99.7 |
| IQ2_XXS | 2.06 | 3.3 GB | 5.0 GB | Runs comfortably | 66K | 118 |
| IQ1_M | 1.75 | 2.9 GB | 4.6 GB | Runs comfortably | 73K | 131 |