← AI Hashrate 中文

RTX A6000 48GB vs V100 32GB SXM2 for Local LLMs

The V100 32GB SXM2 has about 17% more memory bandwidth (900 vs 768 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 17% faster on the V100 32GB SXM2. The RTX A6000 48GB fits more of the catalog at Q4/8K — 46 vs 43 of 73 models (48 vs 32 GB VRAM). Largest model that fits the RTX A6000 48GB but not the V100 32GB SXM2: Qwen3-Next-80B-A3B-Instruct (81.3B). MSRP is $4,500 for the RTX A6000 48GB vs $10,000 for the V100 32GB SXM2. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. Not retail-shoppable: V100 32GB SXM2 (datacenter-class; cloud rental is the realistic way to use it). In the Q4/8K head-to-head below (57 shared models), the V100 32GB SXM2 posts the higher tok/s in 44 rows vs 13. Methodology.

Spec comparison

RTX A6000 48GBV100 32GB SXM2
VendorNVIDIANVIDIA
VRAM48 GB32 GB
Memory typeGDDR6HBM2
Memory bandwidth768 GB/s900 GB/s
TDP300 W300 W
MSRP (list)$4,500$10,000
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 57 shared models

ModelParams (B)RTX A6000 48GB tok/sRTX A6000 48GB fitsV100 32GB SXM2 tok/sV100 32GB SXM2 fitsWinner
MiniCPM5-1B1.08753.6 est.Yes895.6 est.YesV100 32GB SXM2
Qwen3.5-2B2.0407.0 est.Yes483.6 est.YesV100 32GB SXM2
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx2.3353.9 est.Yes420.5 est.YesV100 32GB SXM2
Qwen3.6-35B-A3B36.0271.3 est.Yes322.4 est.YesV100 32GB SXM2
Qwen3.5-35B-A3B34.7271.3 est.Yes322.4 est.YesV100 32GB SXM2
GLM-4.7-Flash31.2271.3 est.Yes322.4 est.YesV100 32GB SXM2
Qwen3-30B-A3B-Instruct-250730.5271.3 est.Yes322.4 est.YesV100 32GB SXM2
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF1630.0271.3 est.Yes322.4 est.YesV100 32GB SXM2
Ornith-1.0-35B35.0271.3 est.Yes322.4 est.YesV100 32GB SXM2
qwen35b-a3b-fable-sft-abliterated34.7271.3 est.Yes322.4 est.YesV100 32GB SXM2
Llama-3.2-3B-Instruct3.2254.4 est.Yes302.3 est.YesV100 32GB SXM2
Qwen3-Coder-Next79.7271.3 est.Yes268.3 est.NoRTX A6000 48GB
Qwen3-Next-80B-A3B-Instruct81.3271.3 est.Yes258.2 est.NoRTX A6000 48GB
gpt-oss-20b21.5226.1 est.Yes268.7 est.YesV100 32GB SXM2
gemma-4-26B-A4B-it25.2214.2 est.Yes254.5 est.YesV100 32GB SXM2

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 A6000 48GBV100 32GB SXM2
Catalog models that fit46 of 7343 of 73
Largest exclusive fitQwen3-Next-80B-A3B-Instruct (81.3B)

RTX A6000 48GB details → · V100 32GB SXM2 details →