RTX 3090 Ti 24GB vs V100 16GB PCIe for Local LLMs

The RTX 3090 Ti 24GB has about 12% more memory bandwidth (1,008 vs 900 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 12% faster on the RTX 3090 Ti 24GB. The RTX 3090 Ti 24GB fits more of the catalog at Q4/8K — 41 vs 25 of 73 models (24 vs 16 GB VRAM). Largest model that fits the RTX 3090 Ti 24GB but not the V100 16GB PCIe: Qwen3.6-35B-A3B (36.0B). MSRP is $1,999 for the RTX 3090 Ti 24GB 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 (49 shared models), the RTX 3090 Ti 24GB posts the higher tok/s in 43 rows vs 6. Methodology.

Spec comparison

	RTX 3090 Ti 24GB	V100 16GB PCIe
Vendor	NVIDIA	NVIDIA
VRAM	24 GB	16 GB
Memory type	GDDR6X	HBM2
Memory bandwidth	1,008 GB/s	900 GB/s
TDP	450 W	250 W
MSRP (list)	$1,999	$8,000
Buy	Search 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 49 shared models

Model	Params (B)	RTX 3090 Ti 24GB tok/s	RTX 3090 Ti 24GB fits	V100 16GB PCIe tok/s	V100 16GB PCIe fits	Winner
MiniCPM5-1B	1.08	1003.1 est.	Yes	895.6 est.	Yes	RTX 3090 Ti 24GB
Qwen3.5-2B	2.0	541.7 est.	Yes	483.6 est.	Yes	RTX 3090 Ti 24GB
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	471.0 est.	Yes	420.5 est.	Yes	RTX 3090 Ti 24GB
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	361.1 est.	Yes	434.0 est.	No	V100 16GB PCIe
Qwen3-30B-A3B-Instruct-2507	30.5	361.1 est.	Yes	405.4 est.	No	V100 16GB PCIe
Qwen3.6-35B-A3B	36.0	361.1 est.	Yes	316.9 est.	No	RTX 3090 Ti 24GB
Qwen3.5-35B-A3B	34.7	361.1 est.	Yes	339.8 est.	No	RTX 3090 Ti 24GB
GLM-4.7-Flash	31.2	361.1 est.	Yes	313.0 est.	No	RTX 3090 Ti 24GB
Ornith-1.0-35B	35.0	361.1 est.	Yes	334.2 est.	No	RTX 3090 Ti 24GB
qwen35b-a3b-fable-sft-abliterated	34.7	361.1 est.	Yes	325.2 est.	No	RTX 3090 Ti 24GB
Llama-3.2-3B-Instruct	3.2	338.5 est.	Yes	302.3 est.	Yes	RTX 3090 Ti 24GB
gpt-oss-20b	21.5	300.9 est.	Yes	268.7 est.	Yes	RTX 3090 Ti 24GB
gemma-4-26B-A4B-it	25.2	285.1 est.	Yes	254.5 est.	Yes	RTX 3090 Ti 24GB
Qwen3-4B-Instruct-2507	4.0	270.8 est.	Yes	241.8 est.	Yes	RTX 3090 Ti 24GB
NVIDIA-Nemotron-3-Nano-4B-BF16	4.0	270.8 est.	Yes	241.8 est.	Yes	RTX 3090 Ti 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 Ti 24GB	V100 16GB PCIe
Catalog models that fit	41 of 73	25 of 73
Largest exclusive fit	Qwen3.6-35B-A3B (36.0B)	—

RTX 3090 Ti 24GB details → · V100 16GB PCIe details →