Nvidia A4000 vs Nvidia H100

The H100 is newer, with 5x the memory and 7.5x the bandwidth. The A4000 is typically cheaper.

At a glance

Memory

5x more on H100

A4000: 16 GB

H100: 80 GB

Memory bandwidth

7.5x higher on H100

A4000: 448 GB/s

H100: 3,350 GB/s

Cheapest on-demand

96% cheaper on A4000

A4000: $0.07 /GPU/hr

H100: $1.73 /GPU/hr

Price comparison

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

Billing	A4000	H100	Difference
On-demand	$0.20 /GPU/hr from $0.07	$3.38 /GPU/hr from $1.73	94% cheaper on A4000
Reserved	$0.18 /GPU/hr from $0.06 (3 mo)	$3.14 /GPU/hr from $1.53 (1 mo)	94% cheaper on A4000
Spot	-- /GPU/hr from $0.06	$2.45 /GPU/hr from $0.67

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. No median for spot on the A4000: we only show one when at least 3 providers list the GPU on that billing type. Verify before provisioning. More on how we price.

Price history

A4000 vs H100 price history

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Specs comparison

Spec	A4000	H100	Difference
Memory
Capacity	16 GB GDDR6	80 GB HBM3 94 GB on NVL	5x more on H100
Bandwidth	448 GB/s	3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL	7.5x higher on H100
Bus width	256-bit	5,120-bit	20x wider on H100
Compute · dense
FP8	Not supported	1,979 TFLOPS	H100 only
FP16 / BF16	76.7 TFLOPS	989.5 TFLOPS	13x higher on H100
INT8		1,979 TOPS
FP32	19.2 TFLOPS	67 TFLOPS	3.5x higher on H100
FP64	Not supported	34 TFLOPS	H100 only
Platform
Interconnect	PCIe only	NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL	NVLink on H100 only
Board power	140 W	700 W	5x higher on H100
Released	Q2 2021	Q3 2022	17 months newer on H100
Precision
Supported formats	FP16BF16TF32FP32INT4INT8	FP8FP16BF16TF32FP32FP64INT8

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

Compute: Vendor peak figures, dense and per GPU. Not measured throughput.

Source: A4000 datasheet and H100 datasheet.

Where to rent them

Providers that rent these GPUs, with each one's cheapest rate. Providers that rent both are listed first.

Provider	A4000	H100
Vast.ai USA	$0.07 /GPU/hr On-demand	$1.73 /GPU/hr On-demand SXM
GPU.ai UAE	$0.09 /GPU/hr On-demand	$2.48 /GPU/hr On-demand PCIe
Hyperstack UK	$0.15 /GPU/hr On-demand	$2.50 /GPU/hr On-demand PCIe
Spheron Singapore	$0.16 /GPU/hr On-demand PCIe	--
Sesterce France	$0.17 /GPU/hr On-demand	--
Runcrate USA	$0.20 /GPU/hr On-demand PCIe	--
EmpirioLabs AI USA	$0.25 /GPU/hr On-demand	--
TensorDock Distributed	$0.21 /GPU/hr On-demand	--
HyperAI Hong Kong	--	$1.80 /GPU/hr On-demand
Lium Distributed	--	$1.90 /GPU/hr On-demand SXM
Runpod USA	--	$2.59 /GPU/hr On-demand NVL
Jarvislabs India	--	$2.69 /GPU/hr On-demand SXM
Massed Compute USA	--	$2.73 /GPU/hr On-demand