Compare XPUs

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

Select XPUs to Compare

Clear all (4)Jump to results

Filter by Vendor

Showing 72 XPUs • 4 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 A100 SXM	NVIDIA L40S	Tenstorrent Wormhole	AWS Inferentia2
Architecture	Ampere	Ada Lovelace	Tensix Core	Inferentia Gen2
Form Factor	SXM	PCIe	—	—
VRAM	80 GB	48 GB	24 GB	32 GB
Memory Bandwidth	2,039 GB/s	864 GB/s	—	—
TFLOPs (FP32)	19.5	91.6	—	—
TFLOPs (FP16)	312	733	—	—
TFLOPs	312	733	364	190
TFLOPs (FP8)	—	1,466	—	—
TDP	400 W	350 W	160 W	150 W
Launch Date	May 2020	Oct 2023	Oct 2023	Nov 2022

Efficiency Metrics

Metric	A100 SXM	L40S	Wormhole	Inferentia2
TFLOPs per Watt (FP32-eq)	0.39	1.05	1.14	0.63
Memory Bandwidth per GB	25.5 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 A100 SXM

NVIDIA L40S

Compute (FP32-eq)

0.43x

L40S is 2.35x faster

FP32 Compute

0.21x

L40S is 4.70x faster

VRAM

1.67x

Need 1.67x L40S

Memory Bandwidth

2.36x

Need 2.36x L40S

Tenstorrent Wormhole

Compute (FP32-eq)

0.86x

Wormhole is 1.17x faster

VRAM

3.33x

Need 3.33x Wormhole

AWS Inferentia2

Compute (FP32-eq)

1.64x

Need 1.64x Inferentia2

VRAM

2.50x

Need 2.50x Inferentia2

To match 1x NVIDIA L40S

NVIDIA A100 SXM

Compute (FP32-eq)

2.35x

Need 2.35x A100 SXM

FP32 Compute

4.70x

Need 4.70x A100 SXM

VRAM

0.60x

A100 SXM has 1.67x more

Memory Bandwidth

0.42x

A100 SXM has 2.36x more

Tenstorrent Wormhole

Compute (FP32-eq)

2.01x

Need 2.01x Wormhole

VRAM

2.00x

Need 2.00x Wormhole

AWS Inferentia2

Compute (FP32-eq)

3.86x

Need 3.86x Inferentia2

VRAM

1.50x

Need 1.50x Inferentia2

To match 1x Tenstorrent Wormhole

NVIDIA A100 SXM

Compute (FP32-eq)

1.17x

Need 1.17x A100 SXM

VRAM

0.30x

A100 SXM has 3.33x more

NVIDIA L40S

Compute (FP32-eq)

0.50x

L40S is 2.01x faster

VRAM

0.50x

L40S has 2.00x more

AWS Inferentia2

Compute (FP32-eq)

1.92x

Need 1.92x Inferentia2

VRAM

0.75x

Inferentia2 has 1.33x more

To match 1x AWS Inferentia2

NVIDIA A100 SXM

Compute (FP32-eq)

0.61x

A100 SXM is 1.64x faster

VRAM

0.40x

A100 SXM has 2.50x more

NVIDIA L40S

Compute (FP32-eq)

0.26x

L40S is 3.86x faster

VRAM

0.67x

L40S has 1.50x more

Tenstorrent Wormhole

Compute (FP32-eq)

0.52x

Wormhole is 1.92x faster

VRAM

1.33x

Need 1.33x Wormhole

Pricing

Price Type	A100 SXM	L40S	Wormhole	Inferentia2
CAPEX (Street Price)	$15,000	$10,000	—	—
OPEX (per hour)	$4.05/hr	$1.50/hr	—	—
Price per TFLOPs (FP32-eq)	$96	$27	—	—