Compare XPUs

Select up to 5 XPUs to compare side-by-side

Select XPUs to Compare

Clear all (3)Jump to results

Filter by Vendor

Showing 72 XPUs • 3 selected

Alibaba

Hanguang 800

AMD

MI100

23.1 TFLOPs

AMD

MI210

181 TFLOPs

AMD

MI250X

383 TFLOPs

AMD

MI300A

980.6 TFLOPs

AMD

MI300X

1,307 TFLOPs

AMD

MI325X

1,400 TFLOPs

AMD

MI350X

2,100 TFLOPs

AMD

Radeon PRO W7900

122 TFLOPs

AMD

Radeon RX 7900 XT

104 TFLOPs

AMD

Radeon RX 7900 XTX

122 TFLOPs

AWS

Inferentia2

190 TFLOPs

AWS

Trainium

190 TFLOPs

AWS

Trainium2

680 TFLOPs

Baidu

Kunlun II

Biren Technology

BR100

Cambricon

MLU370

256 TFLOPs

Cerebras

WSE-3

Enflame Technology

CloudBlazer T20

Etched

Sohu

10,000 TFLOPs

FuriosaAI

Warboy

Google

TPU v4

275 TFLOPs

Google

TPU v5e

197 TFLOPs

Google

TPU v5p

459 TFLOPs

Google

TPU v6e (Trillium)

918 TFLOPs

Graphcore

Bow IPU

Graphcore

IPU-M2000

Groq

LPU Inference Engine

Huawei

Ascend 910B

Iluvatar CoreX

BI-V150

300 TFLOPs

Intel

Data Center GPU Max 1100

177 TFLOPs

Intel

Data Center GPU Max 1550

419 TFLOPs

Intel Habana

Gaudi 2

432 TFLOPs

Intel Habana

Gaudi 3

1,835 TFLOPs

Multi-Metric Comparison

Relative performance across 5 key metrics (normalized to 100 = best in comparison)

Compute Performance (BF16)

Memory Capacity

Power Consumption

Power Efficiency

Specifications

Specification	NVIDIA H200 SXM	AWS Trainium2	NVIDIA RTX 5000 Ada Generation
Architecture	Hopper	Trainium Gen2	Ada Lovelace
Form Factor	SXM	—	PCIe
VRAM	141 GB	—	32 GB
Memory Bandwidth	4,800 GB/s	—	576 GB/s
TFLOPs (FP32)	67	—	65.3
TFLOPs (FP16)	1,979	—	130.6
TFLOPs	1,979	680	130.6
TFLOPs (FP8)	3,958	—	—
TDP	700 W	300 W	250 W
Launch Date	Nov 2023	Nov 2023	Aug 2023

Efficiency Metrics

Metric	H200 SXM	Trainium2	RTX 5000 Ada Generation
TFLOPs per Watt (FP32-eq)	1.41	1.13	0.26
Memory Bandwidth per GB	34.0 GB/s	—	18.0 GB/s

Performance Equivalence

How many units of each GPU are needed to match the performance of the others?

To match 1x NVIDIA H200 SXM

AWS Trainium2

Compute (FP32-eq)

2.91x

Need 2.91x Trainium2

NVIDIA RTX 5000 Ada Generation

Compute (FP32-eq)

15.15x

Need 15.15x RTX 5000 Ada Generation

FP32 Compute

1.03x

Need 1.03x RTX 5000 Ada Generation

VRAM

4.41x

Need 4.41x RTX 5000 Ada Generation

Memory Bandwidth

8.33x

Need 8.33x RTX 5000 Ada Generation

To match 1x AWS Trainium2

NVIDIA H200 SXM

Compute (FP32-eq)

0.34x

H200 SXM is 2.91x faster

NVIDIA RTX 5000 Ada Generation

Compute (FP32-eq)

5.21x

Need 5.21x RTX 5000 Ada Generation

To match 1x NVIDIA RTX 5000 Ada Generation

NVIDIA H200 SXM

Compute (FP32-eq)

0.07x

H200 SXM is 15.15x faster

FP32 Compute

0.97x

H200 SXM is 1.03x faster

VRAM

0.23x

H200 SXM has 4.41x more

Memory Bandwidth

0.12x

H200 SXM has 8.33x more

AWS Trainium2

Compute (FP32-eq)

0.19x

Trainium2 is 5.21x faster

Pricing

Price Type	H200 SXM	Trainium2	RTX 5000 Ada Generation
CAPEX (Street Price)	$40,000	—	—
OPEX (per hour)	—	$1.34/hr	—
Price per TFLOPs (FP32-eq)	$40	—	—