Gemma on NVIDIA Ampere

Can I run Gemma 3 4B on an RTX 3080 10GB?

Yes. Gemma 3 4B at Q4_K_M uses 3.6 GB of the 8.6 GB available on a RTX 3080 10GB, and runs at about 171 tokens per second. You can push the context to 128K.

Runs comfortably

3.6 GB of 8.6 GB · 42%
09 GB
Weights 2.5 GB
KV cache 0.3 GB
Runtime overhead 0.8 GB

Gemma 3 4B at Q4_K_M leaves 5.0 GB spare on a RTX 3080 10GB. There is room to raise the context length or move up a quantisation level.

Generation171tok/s
Prompt processing2930tok/s
Max context128Ktokens
KV per 1K tokens0GB

Every quantisation of Gemma 3 4B on a RTX 3080 10GB

Highlighted row is the highest quality that still fits at 8K context.

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 8.0 GB 9.1 GB Won't fit — 29.3
INT8 / W8A8 8.50 4.1 GB 5.2 GB Runs comfortably 128K 108
Q8_0 (GGUF) 8.50 4.1 GB 5.2 GB Runs comfortably 128K 108
FP8 (E4M3) 8.00 4.0 GB 5.1 GB Runs comfortably 128K 111
Q6_K 6.56 3.3 GB 4.4 GB Runs comfortably 128K 133
AWQ 4-bit 4.25 3.0 GB 4.1 GB Runs comfortably 128K 142
GPTQ 4-bit 4.25 3.0 GB 4.1 GB Runs comfortably 128K 142
MXFP4 4.25 3.0 GB 4.1 GB Runs comfortably 128K 142
Q5_K_M 5.67 2.8 GB 3.9 GB Runs comfortably 128K 152
Q5_K_S 5.52 2.8 GB 3.9 GB Runs comfortably 128K 155
Q4_K_M 4.85 2.5 GB 3.6 GB Runs comfortably 128K 171
Q4_K_S 4.58 2.4 GB 3.5 GB Runs comfortably 128K 179
Q4_0 4.55 2.4 GB 3.4 GB Runs comfortably 128K 179
IQ4_XS 4.25 2.2 GB 3.3 GB Runs comfortably 128K 188
Q3_K_M 3.91 2.1 GB 3.2 GB Runs comfortably 128K 200
IQ3_M 3.70 2.0 GB 3.1 GB Runs comfortably 128K 207
IQ3_XXS 3.06 1.7 GB 2.8 GB Runs comfortably 128K 235
Q2_K 2.63 1.5 GB 2.6 GB Runs comfortably 128K 258
IQ2_XXS 2.06 1.3 GB 2.4 GB Runs comfortably 128K 297
IQ1_M 1.75 1.2 GB 2.3 GB Runs comfortably 128K 324

Also worth checking

Gemma 3 4B on every GPU
Full hardware tableEverything that runs on a RTX 3080 10GB
Full model tableWhy the KV cache is the part that bites
Guide