GeForce RTX 3060 12 GB vs Quadro RTX 6000 — Specs & Performance Compared

Verdict: On llama2:7b, the Quadro RTX 6000 generates 102.8 tokens/s (per llama.cpp GitHub (CUDA performance scoreboard, Discussion #15013)) against 75.9 for the GeForce RTX 3060 12 GB (per llama.cpp GitHub Discussion #10879), best community run on file for each.

Side-by-side GPU comparison of the GeForce RTX 3060 12 GB and the Quadro RTX 6000. Specs, synthetic benchmarks, AI inference tok/s, and Amazon availability below; each column header links to that part's benchmark page.

NVIDIA · Ampere

GeForce RTX 3060 12 GB

12 GB VRAM$329 MSRP

Full GeForce RTX 3060 12 GB benchmarks →

NVIDIA · Turing

Quadro RTX 6000

24 GB VRAM2018

Full Quadro RTX 6000 benchmarks →

*Price sourced from Amazon.com. Price and availability subject to change.

Where to buy

A listing for each part; the current price and stock are on the retailer's page.

Specs side-by-side

Spec	GeForce RTX 3060 12 GB	Quadro RTX 6000
vram gb	12	24

Synthetic benchmarks

Each row says whether a higher or a lower number is better (ranks and times are lower-is-better). Scores come from 3DMark, WCCFTech, TopCPU, UL Benchmarks, Geekbench Browser and PassMark; each row links its sources, and where they differ the cell shows their median. Empty cells mean we don't yet have that benchmark on file for the SKU.

Benchmark	GeForce RTX 3060 12 GB	Quadro RTX 6000
3DMark Fire Strike (higher is better) Source: 3DMark	21,653 points	—
3DMark Fire Strike Ultra (higher is better) Source: 3DMark	—	8,190 points
3DMark Fire Strike Ultra Graphics (higher is better) Source: 3DMark	—	8,245 points
3DMark Port Royal (higher is better) Source: 3DMark, WCCFTech	5,168 points	—
3DMark Speed Way (higher is better) Source: 3DMark, TopCPU	2,294 points	—
3DMark Steel Nomad (higher is better) Source: UL Benchmarks	1,974 points	—
3DMark Time Spy (higher is better) Source: TopCPU, 3DMark	8,718 points	13,758 points
Geekbench 6 OpenCL (higher is better) Source: Geekbench Browser	82,071 points	—
Geekbench 6 Vulkan (higher is better) Source: Geekbench Browser	91,383 points	—
PassMark G2D Mark (higher is better) Source: PassMark	—	754 pts
PassMark G3D Mark (higher is better) Source: PassMark	—	17,828 pts
PassMark GPU Compute (higher is better) Source: PassMark	—	9,044 Ops/Sec

AI inference (tokens/sec)

Top community-reported tokens-per-second results for popular LLMs whose weights fit the card's memory, each linked to its source. See the per-SKU benchmark page for the full set including quantization and prompt length.

GeForce RTX 3060 12 GB

llama3.2:1b: 211.9 tok/s (TYO Lab blog)
llama3.2:1b: 211.9 tok/s (TyoLab)
llama3.2:1b: 185.0 tok/s (LocalScore (Mozilla Builders))

All GeForce RTX 3060 12 GB AI benchmarks →

Quadro RTX 6000

llama3.2:1b-instruct: 253.0 tok/s (LocalScore (Mozilla llamafile benchmark))
llama2:7b: 102.8 tok/s (llama.cpp GitHub (CUDA performance scoreboard, Discussion #15013))
llama2:7b: 102.8 tok/s (llama.cpp GitHub (CUDA performance scoreboard))

All Quadro RTX 6000 AI benchmarks →

Quick answers

Which is better for running local LLMs?

On llama2:7b, the Quadro RTX 6000 generates 102.8 tokens/s (per llama.cpp GitHub (CUDA performance scoreboard, Discussion #15013)) against 75.9 for the GeForce RTX 3060 12 GB (per llama.cpp GitHub Discussion #10879), best community run on file for each.