AMD MI350X vs Nvidia A40

Two data center GPUs, one CDNA 4 and one Ampere: 288 GB HBM3e against 48 GB GDDR6, 11x the memory bandwidth on the MI350X, and Infinity Fabric 1,075 GB/s on the MI350X where the A40 is PCIe only.

At a glance

Memory

MI350X: 288 GB
A40: 48 GB

6x more on MI350X

Memory bandwidth

MI350X: 8,000 GB/s
A40: 696 GB/s

11x higher on MI350X

Cheapest on-demand

MI350X: --
A40: $0.49 /GPU/hr

MI350X is not sold on demand

Price comparison

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

Billing	MI350X	A40
On-demand	No on-demand listings	$0.77 /GPU/hr from $0.49
Reserved	-- /GPU/hr from $4.76 (12 mo)	-- /GPU/hr from $0.61 (24 mo)
Spot	-- /GPU/hr from $4.00	-- /GPU/hr from $0.31

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 MI350X and reserved and spot on the A40: 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

MI350X vs A40 price history

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Specs comparison

Spec	MI350X	A40	Difference
Memory
Capacity	288 GB HBM3e	48 GB GDDR6	6x more on MI350X
Bandwidth	8,000 GB/s	696 GB/s	11x higher on MI350X
Bus width	8,192-bit	384-bit	21x wider on MI350X
Compute · dense
FP4	9,227.5 TFLOPS	Not supported	MI350X only
FP8	4,614 TFLOPS	Not supported	MI350X only
FP16 / BF16	2,309.6 TFLOPS	149.7 TFLOPS	15x higher on MI350X
INT8	4,614 TOPS	299.3 TOPS	15x higher on MI350X
FP32	144.2 TFLOPS	37.4 TFLOPS	3.9x higher on MI350X
FP64	72.1 TFLOPS	Not supported	MI350X only
Platform
Interconnect	Infinity Fabric 1,075 GB/s	PCIe only	Infinity Fabric on MI350X only
Board power	1,000 W	300 W	3.3x higher on MI350X
Released	Q2 2025	Q4 2020	5 years newer on MI350X
Precision
Supported formats	FP4FP6FP8FP16BF16FP32FP64INT8	FP16BF16TF32FP32INT4INT8

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

Source: MI350X datasheet and A40 datasheet.

Where to rent them

Providers listing these GPUs, those with both first, so you can switch without moving clouds.

Provider	MI350X /GPU/hr	A40 /GPU/hr
DigitalOcean	Reserved (12mo) $4.76	--
GPU.ai	--	On-Demand $0.49
Runpod	--	On-Demand $0.49
Vast.ai	--	On-Demand $1.00
Exoscale	--	On-Demand $1.05
Database Mart	--	Reserved (24mo) $0.61
Utho	--	Reserved (36mo) $0.64
Sesterce	--	On-Demand $1.21
EmpirioLabs AI	--	On-Demand $1.49