Nvidia H20 vs Nvidia V100

The H20 is newer, with 3x the memory and 4.4x the bandwidth.

At a glance

Memory

3x more on H20

H20: 96 GB

V100: 32 GB

Memory bandwidth

4.4x higher on H20

H20: 4,000 GB/s

V100: 900 GB/s

Cheapest on-demand

H20: --

V100: $0.11 /GPU/hr

Price comparison

Median price per GPU per hour by billing type, with the cheapest listing under it.

Billing	H20	V100
On-demand	No on-demand listings	$0.95 /GPU/hr from $0.11
Reserved	No reserved listings	$0.97 /GPU/hr from $0.12 (1 mo)
Spot	No spot listings	$0.40 /GPU/hr from $0.09

Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Medians are across providers, one vote each, so a large catalog does not outweigh a small one. Verify before provisioning. More on how we price.

Price history

H20 vs V100 price history

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Specs comparison

Spec	H20	V100	Difference
Memory
Capacity	96 GB HBM3	32 GB HBM2	3x more on H20
Bandwidth	4,000 GB/s	900 GB/s	4.4x higher on H20
Bus width		4,096-bit
Platform
Interconnect	PCIe only	NVLink 300 GB/s	NVLink on V100 only
Board power	400 W	300 W	33% higher on H20
Released	Q1 2024	Q2 2017	6 years newer on H20
Precision
Supported formats	FP8FP16BF16TF32FP32FP64INT8	FP16FP32FP64INT8

Figures are for each model's headline configuration. Where a form factor carries a different number, it is named under the figure.

Source: V100 datasheet.

Where to rent them

Providers that rent these GPUs, with each one's cheapest rate. No provider we track offers both right now.

Provider	H20	V100
Vast.ai USA	--	$0.11 /GPU/hr On-demand
GPU.ai UAE	--	$0.12 /GPU/hr On-demand
HyperAI Hong Kong	--	$0.14 /GPU/hr On-demand
Runpod USA	--	$0.19 /GPU/hr On-demand PCIe
Verda Finland	--	$0.20 /GPU/hr On-demand
Sesterce France	--	$2.57 /GPU/hr On-demand
Runcrate USA	--	$3.04 /GPU/hr On-demand NVL
Geodd USA	--	$0.16 /GPU/hr On-demand