01.AI on NVIDIA Ampere

Can I run Yi-1.5 34B on an RTX 3080 10GB?

Not at Q4_K_M — it needs 22.2 GB against 8.6 GB available. You would need 3 of these cards.

Won't fit

22.2 GB of 8.6 GB · 259%
023 GB
Weights 19.5 GB
KV cache 1.9 GB
Runtime overhead 0.9 GB
Over the limit 13.6 GB

Short by 13.6 GB. You can run it with 18 of 60 layers on the RTX 3080 10GB and the rest in system RAM, at roughly 2.69 tok/s — usable for batch work, painful for chat. A smaller quantisation or a shorter context is usually the better trade.

Generation2.69tok/s
Prompt processing366tok/s
Max context0tokens
KV per 1K tokens0GB

Every quantisation of Yi-1.5 34B 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 64.1 GB 66.8 GB Won't fit — 0.67
INT8 / W8A8 8.50 33.8 GB 36.6 GB Won't fit — 1.36
Q8_0 (GGUF) 8.50 33.8 GB 36.6 GB Won't fit — 1.36
FP8 (E4M3) 8.00 32.0 GB 34.8 GB Won't fit — 1.43
Q6_K 6.56 26.3 GB 29.0 GB Won't fit — 1.83
Q5_K_M 5.67 22.7 GB 25.5 GB Won't fit — 2.19
Q5_K_S 5.52 22.1 GB 24.9 GB Won't fit — 2.25
Q4_K_M 4.85 19.5 GB 22.2 GB Won't fit — 2.69
Q4_K_S 4.58 18.4 GB 21.2 GB Won't fit — 2.90
Q4_0 4.55 18.3 GB 21.1 GB Won't fit — 2.92
AWQ 4-bit 4.25 18.3 GB 21.0 GB Won't fit — 2.93
GPTQ 4-bit 4.25 18.3 GB 21.0 GB Won't fit — 2.93
MXFP4 4.25 18.3 GB 21.0 GB Won't fit — 2.93
IQ4_XS 4.25 17.2 GB 19.9 GB Won't fit — 3.18
Q3_K_M 3.91 15.8 GB 18.6 GB Won't fit — 3.58
IQ3_M 3.70 15.0 GB 17.8 GB Won't fit — 3.86
IQ3_XXS 3.06 12.5 GB 15.3 GB Won't fit — 5.18
Q2_K 2.63 10.8 GB 13.6 GB Won't fit — 6.62
IQ2_XXS 2.06 8.6 GB 11.4 GB Won't fit — 10.7
IQ1_M 1.75 7.4 GB 10.2 GB Won't fit 1K 16.8

Also worth checking

Yi-1.5 34B 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