Home
Compare GPUs
MI250 vs MI355X

AMD MI250 vs AMD MI355X

Two AMD data center GPUs, one CDNA2 and one CDNA 4: 128 GB HBM2e against 288 GB HBM3e and 2.5x the memory bandwidth on the MI355X.

At a glance

Updated 6 minutes ago

AMD MI250

Memory: 128 GB; HBM2e
Memory bandwidth: 3.20 TB/s
From: --; Not sold on demand · 1 provider

AMD MI355X

Memory: 288 GB; HBM3e
Memory bandwidth: 8.00 TB/s
From: $8.60 /GPU/hr; 4 providers

Price comparison

Cheapest listed price per GPU per hour by billing type, and the median across providers.

Billing	MI250	MI355X	Difference
On-demand	Not sold on demand	$8.60
Spot	No spot listings	$4.50
Reserved	$1.30	No reserved listings
Median	$1.50	$8.60	5.7x more expensive on MI355X

Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.

Price history

MI250 vs MI355X price history

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Specs comparison

Spec	MI250	MI355X	Difference
Memory
Capacity	128 GB HBM2e	288 GB HBM3e	2.2x more on MI355X
Bandwidth	3,200 GB/s	8,000 GB/s	2.5x higher on MI355X
Bus width	8,192-bit	8,192-bit	Identical
Compute · dense
FP4	Not supported	10,066.3 TFLOPS	MI355X only
FP8	Not supported	5,033.2 TFLOPS	MI355X only
FP16 / BF16	362.1 TFLOPS	2,516.6 TFLOPS	7x higher on MI355X
INT8	362.1 TOPS	5,033.2 TOPS	14x higher on MI355X
FP32	90.5 TFLOPS	157.3 TFLOPS	1.7x higher on MI355X
FP64	45.3 TFLOPS	78.6 TFLOPS	1.7x higher on MI355X
Platform
Interconnect	Infinity Fabric 400 GB/s	Infinity Fabric 1,075 GB/s	2.7x higher on MI355X
Board power	500 W	1,400 W	2.8x higher on MI355X
Released	Q4 2021	Q2 2025	4 years newer on MI355X
Precision
Supported formats	FP16BF16FP32FP64INT4INT8	FP4FP6FP8FP16BF16FP32FP64INT8

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

Source: MI250 datasheet and MI355X datasheet.

Where to rent them

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

Provider	MI250 /GPU/hr	MI355X /GPU/hr
Cirrascale	Reserved (12mo) $1.30	--
Oracle Cloud	--	On-Demand $8.60
DigitalOcean	--	Spot $4.50
TensorWave	--	Custom $2.95
Vultr	--	Spot $2.59

Built for agents: CLI, SDK and MCP server. Your agent rents and runs GPUs itself.

A permissionless marketplace. 1000+ nodes from independent providers worldwide.

Hard to find nodes available; B200 and B300 Blackwell nodes listed now.

Consumer + DC grade cards for cheap. H100s from $1/hr, RTX 5090s from $0.35/hr.

Similar GPUs

Other comparisons in the same class.

Nvidia A100

80 GB HBM2e · 40% less bandwidth than the MI250

From $0.89 /GPU/hr Compare
AMD MI300X

192 GB HBM3 · 66% more bandwidth than the MI250

From $2.59 /GPU/hr Compare
Nvidia V100

32 GB HBM2 · 72% less bandwidth than the MI250

From $0.14 /GPU/hr Compare
Nvidia H100

80 GB HBM3 · 58% less bandwidth than the MI355X

From $1.50 /GPU/hr Compare
Nvidia B200

180 GB HBM3e · released Q1 2024

From $3.75 /GPU/hr Compare

Browse all GPUs

Our data for AMD MI250 was last updated on Sept. 7, 2026, and for AMD MI355X on Sept. 7, 2026.