Nvidia GB200 vs Nvidia H100

The GB200 is newer, with 2.3x the memory and 2.4x the bandwidth. The H100 is typically cheaper.

At a glance

Memory

2.3x more on GB200

GB200: 186 GB

H100: 80 GB

Memory bandwidth

2.4x higher on GB200

GB200: 8.00 TB/s

H100: 3.35 TB/s

Cheapest on-demand

88% cheaper on H100

GB200: $10.50 /GPU/hr

H100: $1.30 /GPU/hr

Price comparison

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

Billing	GB200	H100	Difference
On-demand	$16.00 /GPU/hr from $10.50	$3.38 /GPU/hr from $1.30	79% cheaper on H100
Reserved	-- /GPU/hr from $10.58	$3.09 /GPU/hr from $1.53 (1 mo)
Spot	-- /GPU/hr from $27.04	$2.23 /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 reserved and spot on the GB200: 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

GB200 vs H100 price history

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Specs comparison

Spec	GB200	H100	Difference
Memory
Capacity	186 GB HBM3e	80 GB HBM3 94 GB on NVL	2.3x more on GB200
Bandwidth	8,000 GB/s	3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL	2.4x higher on GB200
Bus width		5,120-bit
Compute · dense
FP4	10,000 TFLOPS	Not supported	GB200 only
FP8	5,000 TFLOPS	1,979 TFLOPS	2.5x higher on GB200
FP16 / BF16	2,500 TFLOPS	989.5 TFLOPS	2.5x higher on GB200
INT8	5,000 TOPS	1,979 TOPS	2.5x higher on GB200
FP32	80 TFLOPS	67 TFLOPS	19% higher on GB200
FP64	40 TFLOPS	34 TFLOPS	18% higher on GB200
Platform
Interconnect	NVLink 1,800 GB/s	NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL	2x higher on GB200
Board power	1,200 W	700 W	71% higher on GB200
Released	Q1 2024	Q3 2022	2 years newer on GB200
Precision
Supported formats	FP4FP6FP8FP16BF16TF32FP32FP64INT8	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: GB200 datasheet and H100 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	GB200	H100
Denvr Canada	On request Custom	--
CoreWeave USA	$10.50 /GPU/hr On-demand NVL	--
Oracle Cloud USA	$16.00 /GPU/hr On-demand NVL	--
Azure USA	$27.04 /GPU/hr On-demand	--
AWS USA	$10.58 /GPU/hr Reserved	--
CUDO UK	On request Custom NVL	--
Canopy Wave USA	On request Custom NVL	--
Crusoe USA	On request Custom NVL	--
Lium Distributed	--	$1.30 /GPU/hr On-demand SXM
Vast.ai USA	--	$1.73 /GPU/hr On-demand SXM
HyperAI Hong Kong	--	$1.80 /GPU/hr On-demand
Hyperstack UK	--	$2.50 /GPU/hr On-demand PCIe
EmpirioLabs AI USA	--	$2.60 /GPU/hr On-demand
Jarvislabs India	--	$2.69 /GPU/hr On-demand SXM
GPU.ai UAE	--	$2.89 /GPU/hr On-demand PCIe
Runpod USA	--	$2.89 /GPU/hr On-demand PCIe