← AI Hashrate 中文

RTX 3080 10GB vs V100 16GB PCIe for Local LLMs

The V100 16GB PCIe has about 18% more memory bandwidth (900 vs 760 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 18% faster on the V100 16GB PCIe. The V100 16GB PCIe fits more of the catalog at Q4/8K — 25 vs 17 of 73 models (16 vs 10 GB VRAM). Largest model that fits the V100 16GB PCIe but not the RTX 3080 10GB: gemma-4-26B-A4B-it (25.2B). MSRP is $699 for the RTX 3080 10GB vs $8,000 for the V100 16GB PCIe. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. Not retail-shoppable: V100 16GB PCIe (datacenter-class; cloud rental is the realistic way to use it). In the Q4/8K head-to-head below (43 shared models), the V100 16GB PCIe posts the higher tok/s in 39 rows vs 4. Methodology.

Spec comparison

RTX 3080 10GBV100 16GB PCIe
VendorNVIDIANVIDIA
VRAM10 GB16 GB
Memory typeGDDR6XHBM2
Memory bandwidth760 GB/s900 GB/s
TDP320 W250 W
MSRP (list)$699$8,000
BuySearch Amazon · eBay · Newegg · B&H Photo

MSRP is the launch list price, not live retail. Buy links are Amazon affiliate links — the price after you click is the live price.

Q4 / 8K head-to-head — top 15 of 43 shared models

ModelParams (B)RTX 3080 10GB tok/sRTX 3080 10GB fitsV100 16GB PCIe tok/sV100 16GB PCIe fitsWinner
MiniCPM5-1B1.08745.8 est.Yes895.6 est.YesV100 16GB PCIe
Qwen3.5-2B2.0402.7 est.Yes483.6 est.YesV100 16GB PCIe
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF1630.0142.9 est.No434.0 est.NoV100 16GB PCIe
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx2.3350.2 est.Yes420.5 est.YesV100 16GB PCIe
Qwen3-30B-A3B-Instruct-250730.5133.5 est.No405.4 est.NoV100 16GB PCIe
Qwen3.5-35B-A3B34.7111.9 est.No339.8 est.NoV100 16GB PCIe
Ornith-1.0-35B35.0110.0 est.No334.2 est.NoV100 16GB PCIe
qwen35b-a3b-fable-sft-abliterated34.7107.1 est.No325.2 est.NoV100 16GB PCIe
Qwen3.6-35B-A3B36.0104.3 est.No316.9 est.NoV100 16GB PCIe
GLM-4.7-Flash31.2103.0 est.No313.0 est.NoV100 16GB PCIe
Llama-3.2-3B-Instruct3.2251.7 est.Yes302.3 est.YesV100 16GB PCIe
gpt-oss-20b21.5225.6 est.No268.7 est.YesV100 16GB PCIe
gemma-4-26B-A4B-it25.2157.2 est.No254.5 est.YesV100 16GB PCIe
Qwen3-4B-Instruct-25074.0201.4 est.Yes241.8 est.YesV100 16GB PCIe
NVIDIA-Nemotron-3-Nano-4B-BF164.0201.4 est.Yes241.8 est.YesV100 16GB PCIe

est. = bandwidth-based estimate; measured = curated real benchmark that overrides the estimate. Fits = weights + KV(8K) + 1 GB overhead ≤ 95% of VRAM.

Fit coverage at Q4 / 8K

RTX 3080 10GBV100 16GB PCIe
Catalog models that fit17 of 7325 of 73
Largest exclusive fitgemma-4-26B-A4B-it (25.2B)

RTX 3080 10GB details → · V100 16GB PCIe details →