Llama on NVIDIA Ampere

Can I run Llama 3.1 70B on an RTX 3080 10GB?

Not at Q4_K_M — it needs 43.4 GB against 8.6 GB available. You would need 6 of these cards.

Won't fit

43.4 GB of 8.6 GB · 504%
045 GB
Weights 40.0 GB
KV cache 2.5 GB
Runtime overhead 0.9 GB
Over the limit 34.8 GB

Short by 34.8 GB. You can run it with 10 of 80 layers on the RTX 3080 10GB and the rest in system RAM, at roughly 1.09 tok/s — usable for batch work, painful for chat. A smaller quantisation or a shorter context is usually the better trade.

Generation1.09tok/s
Prompt processing179tok/s
Max context0tokens
KV per 1K tokens0GB

Every quantisation of Llama 3.1 70B on a RTX 3080 10GB

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

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 131.4 GB 134.8 GB Won't fit — 0.31
INT8 / W8A8 8.50 69.3 GB 72.7 GB Won't fit — 0.61
Q8_0 (GGUF) 8.50 69.3 GB 72.7 GB Won't fit — 0.61
FP8 (E4M3) 8.00 65.7 GB 69.1 GB Won't fit — 0.64
Q6_K 6.56 53.9 GB 57.3 GB Won't fit — 0.79
Q5_K_M 5.67 46.6 GB 50.0 GB Won't fit — 0.92
Q5_K_S 5.52 45.3 GB 48.7 GB Won't fit — 0.96
Q4_K_M 4.85 40.0 GB 43.4 GB Won't fit — 1.09
Q4_K_S 4.58 37.8 GB 41.2 GB Won't fit — 1.17
AWQ 4-bit 4.25 37.8 GB 41.2 GB Won't fit — 1.17
GPTQ 4-bit 4.25 37.8 GB 41.2 GB Won't fit — 1.17
MXFP4 4.25 37.8 GB 41.2 GB Won't fit — 1.17
Q4_0 4.55 37.6 GB 41.0 GB Won't fit — 1.18
IQ4_XS 4.25 35.2 GB 38.6 GB Won't fit — 1.25
Q3_K_M 3.91 32.5 GB 35.9 GB Won't fit — 1.37
IQ3_M 3.70 30.8 GB 34.2 GB Won't fit — 1.46
IQ3_XXS 3.06 25.7 GB 29.1 GB Won't fit — 1.81
Q2_K 2.63 22.3 GB 25.7 GB Won't fit — 2.13
IQ2_XXS 2.06 17.8 GB 21.2 GB Won't fit — 2.85
IQ1_M 1.75 15.3 GB 18.7 GB Won't fit — 3.49

Also worth checking

Llama 3.1 70B 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