AMD MI350P vs Nvidia H100

The MI350P is newer, with 80% more memory and 19% more bandwidth.

At a glance

Memory

80% more on MI350P

MI350P: 144 GB

H100: 80 GB

Memory bandwidth

19% higher on MI350P

MI350P: 4.00 TB/s

H100: 3.35 TB/s

Cheapest on-demand

MI350P: --

H100: $1.80 /GPU/hr

Price comparison

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

Billing	MI350P	H100
On-demand	No on-demand listings	$3.99 /GPU/hr from $1.80
Reserved	No reserved listings	$3.17 /GPU/hr from $1.53 (1 mo)
Spot	No spot listings	$2.46 /GPU/hr from $0.74

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

MI350P vs H100 price history

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Specs comparison

Spec	MI350P	H100	Difference
Memory
Capacity	144 GB HBM3e	80 GB HBM3 94 GB on NVL	80% more on MI350P
Bandwidth	4,000 GB/s	3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL	19% higher on MI350P
Bus width	4,096-bit	5,120-bit	25% wider on H100
Compute · dense
FP4	4,600 TFLOPS	Not supported	MI350P only
FP8	2,300 TFLOPS	1,979 TFLOPS	16% higher on MI350P
FP16 / BF16	1,150 TFLOPS	989.5 TFLOPS	16% higher on MI350P
INT8	2,300 TOPS	1,979 TOPS	16% higher on MI350P
FP32	72 TFLOPS	67 TFLOPS	7% higher on MI350P
FP64	36 TFLOPS	34 TFLOPS	6% higher on MI350P
Platform
Interconnect	PCIe only	NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL	NVLink on H100 only
Board power	600 W	700 W	17% higher on H100
Released	Q2 2026	Q3 2022	4 years newer on MI350P
Precision
Supported formats	FP4FP6FP8FP16BF16FP32FP64INT8	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: MI350P 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	MI350P	H100
HyperAI Hong Kong	--	$1.80 /GPU/hr On-demand
GPU.ai UAE	--	$2.38 /GPU/hr On-demand PCIe
UpCloud Finland	--	$2.50 /GPU/hr On-demand SXM
Spheron Singapore	--	$2.65 /GPU/hr On-demand PCIe
Lium Distributed	--	$2.90 /GPU/hr On-demand SXM
Massed Compute USA	--	$3.11 /GPU/hr On-demand NVL
Scaleway France	--	$3.17 /GPU/hr On-demand PCIe
Thunder Compute USA	--	$3.20 /GPU/hr On-demand SXM