← AI Hashrate 中文

RTX 3090 24GB vs Titan X Pascal 12GB for Local LLMs

The RTX 3090 24GB has about 95% more memory bandwidth (936 vs 480 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 95% faster on the RTX 3090 24GB. The RTX 3090 24GB fits more of the catalog at Q4/8K — 35 vs 18 of 69 models (24 vs 12 GB VRAM). Largest model that fits the RTX 3090 24GB but not the Titan X Pascal 12GB: Qwen3.6-35B-A3B (36.0B). MSRP is $1,499 for the RTX 3090 24GB vs $1,200 for the Titan X Pascal 12GB. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. In the Q4/8K head-to-head below (41 shared models), the RTX 3090 24GB posts the higher tok/s in 37 rows vs 4. Methodology.

Spec comparison

RTX 3090 24GBTitan X Pascal 12GB
VendorNVIDIANVIDIA
VRAM24 GB12 GB
Memory typeGDDR6XGDDR5X
Memory bandwidth936 GB/s480 GB/s
TDP350 W250 W
MSRP (list)$1,499$1,200
BuyMSI GeForce RTX 3090 Ventus 3X 24G OC (affiliate) · More listingsSearch Amazon (affiliate)

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 41 shared models

ModelParams (B)RTX 3090 24GB tok/sRTX 3090 24GB fitsTitan X Pascal 12GB tok/sTitan X Pascal 12GB fitsWinner
MiniCPM5-1B1.08551.5 est.Yes282.8 est.YesRTX 3090 24GB
Qwen3.5-2B2.0297.8 est.Yes152.7 est.YesRTX 3090 24GB
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx2.3259.0 est.Yes132.8 est.YesRTX 3090 24GB
Qwen3.6-35B-A3B36.0198.5 est.Yes95.7 est.NoRTX 3090 24GB
Qwen3.5-35B-A3B34.7198.5 est.Yes102.6 est.NoRTX 3090 24GB
GLM-4.7-Flash31.2198.5 est.Yes118.5 est.NoRTX 3090 24GB
Qwen3-30B-A3B-Instruct-250730.5198.5 est.Yes123.5 est.NoRTX 3090 24GB
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF1630.0198.5 est.Yes135.8 est.NoRTX 3090 24GB
Ornith-1.0-35B35.0198.5 est.Yes96.0 est.NoRTX 3090 24GB
qwen35b-a3b-fable-sft-abliterated34.7198.5 est.Yes97.5 est.NoRTX 3090 24GB
gpt-oss-20b21.5161.8 measuredYes189.3 est.NoTitan X Pascal 12GB
Llama-3.2-3B-Instruct3.2186.1 est.Yes95.5 est.YesRTX 3090 24GB
Qwen3-Coder-Next79.7161.7 est.No20.7 est.NoRTX 3090 24GB
gemma-4-26B-A4B-it25.2156.7 est.Yes135.0 est.NoRTX 3090 24GB
Qwen3-Next-80B-A3B-Instruct81.3155.5 est.No19.9 est.NoRTX 3090 24GB

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 3090 24GBTitan X Pascal 12GB
Catalog models that fit35 of 6918 of 69
Largest exclusive fitQwen3.6-35B-A3B (36.0B)

RTX 3090 24GB details → · Titan X Pascal 12GB details →