H100 80GB SXM5 vs V100 32GB SXM2 for Local LLMs

The H100 80GB SXM5 has about 272% more memory bandwidth (3,350 vs 900 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 272% faster on the H100 80GB SXM5. The H100 80GB SXM5 fits more of the catalog at Q4/8K — 55 vs 43 of 73 models (80 vs 32 GB VRAM). Largest model that fits the H100 80GB SXM5 but not the V100 32GB SXM2: Mistral-Medium-3.5-128B (128.0B). MSRP is $30,000 for the H100 80GB SXM5 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: H100 80GB SXM5, 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 H100 80GB SXM5 posts the higher tok/s in 57 rows vs 0. Methodology.

Spec comparison

	H100 80GB SXM5	V100 32GB SXM2
Vendor	NVIDIA	NVIDIA
VRAM	80 GB	32 GB
Memory type	HBM3	HBM2
Memory bandwidth	3,350 GB/s	900 GB/s
TDP	700 W	300 W
MSRP (list)	$30,000	$10,000
Buy	—	—

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

Model	Params (B)	H100 80GB SXM5 tok/s	H100 80GB SXM5 fits	V100 32GB SXM2 tok/s	V100 32GB SXM2 fits	Winner
MiniCPM5-1B	1.08	3366.4 est.	Yes	895.6 est.	Yes	H100 80GB SXM5
Qwen3.5-2B	2.0	1817.8 est.	Yes	483.6 est.	Yes	H100 80GB SXM5
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	1580.7 est.	Yes	420.5 est.	Yes	H100 80GB SXM5
Qwen3.6-35B-A3B	36.0	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
Qwen3.5-35B-A3B	34.7	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
GLM-4.7-Flash	31.2	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
Qwen3-30B-A3B-Instruct-2507	30.5	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
Qwen3-Coder-Next	79.7	1211.9 est.	Yes	268.3 est.	No	H100 80GB SXM5
Ornith-1.0-35B	35.0	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
Qwen3-Next-80B-A3B-Instruct	81.3	1211.9 est.	Yes	258.2 est.	No	H100 80GB SXM5
qwen35b-a3b-fable-sft-abliterated	34.7	1211.9 est.	Yes	322.4 est.	Yes	H100 80GB SXM5
Llama-3.2-3B-Instruct	3.2	1136.1 est.	Yes	302.3 est.	Yes	H100 80GB SXM5
gpt-oss-20b	21.5	1009.9 est.	Yes	268.7 est.	Yes	H100 80GB SXM5
gemma-4-26B-A4B-it	25.2	956.8 est.	Yes	254.5 est.	Yes	H100 80GB SXM5

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

	H100 80GB SXM5	V100 32GB SXM2
Catalog models that fit	55 of 73	43 of 73
Largest exclusive fit	Mistral-Medium-3.5-128B (128.0B)	—

H100 80GB SXM5 details → · V100 32GB SXM2 details →