Llama on NVIDIA Ampere

Can I run Llama 3.2 1B on an RTX 3080 10GB?

Yes. Llama 3.2 1B at Q4_K_M uses 1.8 GB of the 8.6 GB available on a RTX 3080 10GB, and runs at about 534 tokens per second. You can push the context to 128K.

Runs comfortably

1.8 GB of 8.6 GB · 20%
09 GB
Weights 0.7 GB
KV cache 0.3 GB
Runtime overhead 0.8 GB

Llama 3.2 1B at Q4_K_M leaves 6.8 GB spare on a RTX 3080 10GB. There is room to raise the context length or move up a quantisation level.

Generation534tok/s
Prompt processing10161tok/s
Max context128Ktokens
KV per 1K tokens0GB

Every quantisation of Llama 3.2 1B 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 2.3 GB 3.3 GB Runs comfortably 128K 192
INT8 / W8A8 8.50 1.2 GB 2.2 GB Runs comfortably 128K 355
Q8_0 (GGUF) 8.50 1.2 GB 2.2 GB Runs comfortably 128K 355
FP8 (E4M3) 8.00 1.2 GB 2.2 GB Runs comfortably 128K 359
AWQ 4-bit 4.25 1.0 GB 2.0 GB Runs comfortably 128K 416
GPTQ 4-bit 4.25 1.0 GB 2.0 GB Runs comfortably 128K 416
MXFP4 4.25 1.0 GB 2.0 GB Runs comfortably 128K 416
Q6_K 6.56 0.9 GB 2.0 GB Runs comfortably 128K 426
Q5_K_M 5.67 0.8 GB 1.9 GB Runs comfortably 128K 481
Q5_K_S 5.52 0.8 GB 1.8 GB Runs comfortably 128K 492
Q4_K_M 4.85 0.7 GB 1.8 GB Runs comfortably 128K 534
Q4_K_S 4.58 0.7 GB 1.7 GB Runs comfortably 128K 553
Q4_0 4.55 0.7 GB 1.7 GB Runs comfortably 128K 555
IQ4_XS 4.25 0.7 GB 1.7 GB Runs comfortably 128K 578
Q3_K_M 3.91 0.6 GB 1.6 GB Runs comfortably 128K 607
IQ3_M 3.70 0.6 GB 1.6 GB Runs comfortably 128K 626
IQ3_XXS 3.06 0.5 GB 1.6 GB Runs comfortably 128K 692
Q2_K 2.63 0.5 GB 1.5 GB Runs comfortably 128K 746
IQ2_XXS 2.06 0.4 GB 1.4 GB Runs comfortably 128K 830
IQ1_M 1.75 0.4 GB 1.4 GB Runs comfortably 128K 884

Also worth checking

Llama 3.2 1B 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