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 L40S	NVIDIA A40	Meta MTIA v1	NVIDIA GeForce RTX 4060 Ti
Architecture	Ada Lovelace	Ampere	Meta Training & Inference Accelerator	Ada Lovelace
Form Factor	PCIe	PCIe	—	PCIe
VRAM	48 GB	48 GB	128 GB	8 GB
Memory Bandwidth	864 GB/s	696 GB/s	—	288 GB/s
TFLOPs (FP32)	91.6	37.4	—	22.1
TFLOPs (FP16)	733	150	—	44.2
TFLOPs	733	150	—	44.2
TFLOPs (FP8)	1,466	—	—	—
TDP	350 W	300 W	400 W	160 W
Launch Date	Oct 2023	Oct 2020	May 2023	May 2023

Efficiency Metrics

Metric	L40S	A40	MTIA v1	GeForce RTX 4060 Ti
TFLOPs per Watt (FP32-eq)	1.05	0.25	—	0.14
Memory Bandwidth per GB	18.0 GB/s	14.5 GB/s	—	36.0 GB/s

Performance Equivalence

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

To match 1x NVIDIA L40S

NVIDIA A40

Compute (FP32-eq)

4.89x

Need 4.89x A40

FP32 Compute

2.45x

Need 2.45x A40

VRAM

1.00x

A40 has 1.00x more

Memory Bandwidth

1.24x

Need 1.24x A40

Meta MTIA v1

VRAM

0.38x

MTIA v1 has 2.67x more

NVIDIA GeForce RTX 4060 Ti

Compute (FP32-eq)

16.58x

Need 16.58x GeForce RTX 4060 Ti

FP32 Compute

4.14x

Need 4.14x GeForce RTX 4060 Ti

VRAM

6.00x

Need 6.00x GeForce RTX 4060 Ti

Memory Bandwidth

3.00x

Need 3.00x GeForce RTX 4060 Ti

To match 1x NVIDIA A40

NVIDIA L40S

Compute (FP32-eq)

0.20x

L40S is 4.89x faster

FP32 Compute

0.41x

L40S is 2.45x faster

VRAM

1.00x

L40S has 1.00x more

Memory Bandwidth

0.81x

L40S has 1.24x more

Meta MTIA v1

VRAM

0.38x

MTIA v1 has 2.67x more

NVIDIA GeForce RTX 4060 Ti

Compute (FP32-eq)

3.39x

Need 3.39x GeForce RTX 4060 Ti

FP32 Compute

1.69x

Need 1.69x GeForce RTX 4060 Ti

VRAM

6.00x

Need 6.00x GeForce RTX 4060 Ti

Memory Bandwidth

2.42x

Need 2.42x GeForce RTX 4060 Ti

To match 1x Meta MTIA v1

NVIDIA L40S

VRAM

2.67x

Need 2.67x L40S

NVIDIA A40

VRAM

2.67x

Need 2.67x A40

NVIDIA GeForce RTX 4060 Ti

VRAM

16.00x

Need 16.00x GeForce RTX 4060 Ti

To match 1x NVIDIA GeForce RTX 4060 Ti

NVIDIA L40S

Compute (FP32-eq)

0.06x

L40S is 16.58x faster

FP32 Compute

0.24x

L40S is 4.14x faster

VRAM

0.17x

L40S has 6.00x more

Memory Bandwidth

0.33x

L40S has 3.00x more

NVIDIA A40

Compute (FP32-eq)

0.29x

A40 is 3.39x faster

FP32 Compute

0.59x

A40 is 1.69x faster

VRAM

0.17x

A40 has 6.00x more

Memory Bandwidth

0.41x

A40 has 2.42x more

Meta MTIA v1

VRAM

0.06x

MTIA v1 has 16.00x more

Pricing

Price Type	L40S	A40	MTIA v1	GeForce RTX 4060 Ti
CAPEX (Street Price)	$10,000	—	—	—
OPEX (per hour)	$1.50/hr	—	—	—
Price per TFLOPs (FP32-eq)	$27	—	—	—