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 B200	NVIDIA GeForce RTX 5070 Ti	Huawei Ascend 910B
Architecture	Ada Lovelace	Blackwell	Blackwell	Da Vinci
Form Factor	PCIe	SXM	PCIe	—
VRAM	48 GB	192 GB	16 GB	64 GB
Memory Bandwidth	864 GB/s	8,000 GB/s	896 GB/s	—
TFLOPs (FP32)	91.6	90	44	—
TFLOPs (FP16)	733	2,250	88	640
TFLOPs	733	2,250	88	—
TFLOPs (FP8)	1,466	4,500	—	—
TDP	350 W	1000 W	300 W	400 W
Launch Date	Oct 2023	Mar 2024	Feb 2025	Aug 2023

Efficiency Metrics

Metric	L40S	B200	GeForce RTX 5070 Ti	Ascend 910B
TFLOPs per Watt (FP32-eq)	1.05	1.13	0.15	—
Memory Bandwidth per GB	18.0 GB/s	41.7 GB/s	56.0 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 B200

Compute (FP32-eq)

0.33x

B200 is 3.07x faster

FP32 Compute

1.02x

Need 1.02x B200

VRAM

0.25x

B200 has 4.00x more

Memory Bandwidth

0.11x

B200 has 9.26x more

NVIDIA GeForce RTX 5070 Ti

Compute (FP32-eq)

8.33x

Need 8.33x GeForce RTX 5070 Ti

FP32 Compute

2.08x

Need 2.08x GeForce RTX 5070 Ti

VRAM

3.00x

Need 3.00x GeForce RTX 5070 Ti

Memory Bandwidth

0.96x

GeForce RTX 5070 Ti has 1.04x more

Huawei Ascend 910B

VRAM

0.75x

Ascend 910B has 1.33x more

To match 1x NVIDIA B200

NVIDIA L40S

Compute (FP32-eq)

3.07x

Need 3.07x L40S

FP32 Compute

0.98x

L40S is 1.02x faster

VRAM

4.00x

Need 4.00x L40S

Memory Bandwidth

9.26x

Need 9.26x L40S

NVIDIA GeForce RTX 5070 Ti

Compute (FP32-eq)

25.57x

Need 25.57x GeForce RTX 5070 Ti

FP32 Compute

2.05x

Need 2.05x GeForce RTX 5070 Ti

VRAM

12.00x

Need 12.00x GeForce RTX 5070 Ti

Memory Bandwidth

8.93x

Need 8.93x GeForce RTX 5070 Ti

Huawei Ascend 910B

VRAM

3.00x

Need 3.00x Ascend 910B

To match 1x NVIDIA GeForce RTX 5070 Ti

NVIDIA L40S

Compute (FP32-eq)

0.12x

L40S is 8.33x faster

FP32 Compute

0.48x

L40S is 2.08x faster

VRAM

0.33x

L40S has 3.00x more

Memory Bandwidth

1.04x

Need 1.04x L40S

NVIDIA B200

Compute (FP32-eq)

0.04x

B200 is 25.57x faster

FP32 Compute

0.49x

B200 is 2.05x faster

VRAM

0.08x

B200 has 12.00x more

Memory Bandwidth

0.11x

B200 has 8.93x more

Huawei Ascend 910B

VRAM

0.25x

Ascend 910B has 4.00x more

To match 1x Huawei Ascend 910B

NVIDIA L40S

VRAM

1.33x

Need 1.33x L40S

NVIDIA B200

VRAM

0.33x

B200 has 3.00x more

NVIDIA GeForce RTX 5070 Ti

VRAM

4.00x

Need 4.00x GeForce RTX 5070 Ti

Pricing

Price Type	L40S	B200	GeForce RTX 5070 Ti	Ascend 910B
CAPEX (Street Price)	$10,000	$70,000	—	—
OPEX (per hour)	$1.50/hr	—	—	—
Price per TFLOPs (FP32-eq)	$27	$62	—	—