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	AMD MI350X	Enflame Technology CloudBlazer T20	Qualcomm Cloud AI 100
Architecture	Ada Lovelace	CDNA 4	GCU (AI Graphics Unit)	Qualcomm AI
Form Factor	PCIe	OAM	—	PCIe
VRAM	48 GB	288 GB	32 GB	16 GB
Memory Bandwidth	864 GB/s	—	—	134 GB/s
TFLOPs (FP32)	91.6	—	—	50
TFLOPs (FP16)	733	—	—	400
TFLOPs	733	2,100	—	50
TFLOPs (FP8)	1,466	—	—	—
TDP	350 W	850 W	250 W	75 W
Launch Date	Oct 2023	Jun 2025	Nov 2021	Sep 2020

Efficiency Metrics

Metric	L40S	MI350X	CloudBlazer T20	Cloud AI 100
TFLOPs per Watt (FP32-eq)	1.05	1.24	—	0.67
Memory Bandwidth per GB	18.0 GB/s	—	—	8.4 GB/s

Performance Equivalence

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

To match 1x NVIDIA L40S

AMD MI350X

Compute (FP32-eq)

0.35x

MI350X is 2.86x faster

VRAM

0.17x

MI350X has 6.00x more

Enflame Technology CloudBlazer T20

VRAM

1.50x

Need 1.50x CloudBlazer T20

Qualcomm Cloud AI 100

Compute (FP32-eq)

7.33x

Need 7.33x Cloud AI 100

FP32 Compute

1.83x

Need 1.83x Cloud AI 100

VRAM

3.00x

Need 3.00x Cloud AI 100

Memory Bandwidth

6.45x

Need 6.45x Cloud AI 100

To match 1x AMD MI350X

NVIDIA L40S

Compute (FP32-eq)

2.86x

Need 2.86x L40S

VRAM

6.00x

Need 6.00x L40S

Enflame Technology CloudBlazer T20

VRAM

9.00x

Need 9.00x CloudBlazer T20

Qualcomm Cloud AI 100

Compute (FP32-eq)

21.00x

Need 21.00x Cloud AI 100

VRAM

18.00x

Need 18.00x Cloud AI 100

To match 1x Enflame Technology CloudBlazer T20

NVIDIA L40S

VRAM

0.67x

L40S has 1.50x more

AMD MI350X

VRAM

0.11x

MI350X has 9.00x more

Qualcomm Cloud AI 100

VRAM

2.00x

Need 2.00x Cloud AI 100

To match 1x Qualcomm Cloud AI 100

NVIDIA L40S

Compute (FP32-eq)

0.14x

L40S is 7.33x faster

FP32 Compute

0.55x

L40S is 1.83x faster

VRAM

0.33x

L40S has 3.00x more

Memory Bandwidth

0.16x

L40S has 6.45x more

AMD MI350X

Compute (FP32-eq)

0.05x

MI350X is 21.00x faster

VRAM

0.06x

MI350X has 18.00x more

Enflame Technology CloudBlazer T20

VRAM

0.50x

CloudBlazer T20 has 2.00x more

Pricing

Price Type	L40S	MI350X	CloudBlazer T20	Cloud AI 100
CAPEX (Street Price)	$10,000	—	—	—
OPEX (per hour)	$1.50/hr	—	—	—
Price per TFLOPs (FP32-eq)	$27	—	—	—