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	AMD MI100	NVIDIA B200	AWS Trainium2	NVIDIA GeForce RTX 5070 Ti
Architecture	CDNA	Blackwell	Trainium Gen2	Blackwell
Form Factor	PCIe	SXM	—	PCIe
VRAM	32 GB	192 GB	—	16 GB
Memory Bandwidth	1,228 GB/s	8,000 GB/s	—	896 GB/s
TFLOPs (FP32)	23.1	90	—	44
TFLOPs (FP16)	184.6	2,250	—	88
TFLOPs	23.1	2,250	680	88
TFLOPs (FP8)	—	4,500	—	—
TDP	300 W	1000 W	300 W	300 W
Launch Date	Nov 2020	Mar 2024	Nov 2023	Feb 2025

Efficiency Metrics

Metric	MI100	B200	Trainium2	GeForce RTX 5070 Ti
TFLOPs per Watt (FP32-eq)	0.08	1.13	1.13	0.15
Memory Bandwidth per GB	38.4 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 AMD MI100

NVIDIA B200

Compute (FP32-eq)

0.02x

B200 is 48.70x faster

FP32 Compute

0.26x

B200 is 3.90x faster

VRAM

0.17x

B200 has 6.00x more

Memory Bandwidth

0.15x

B200 has 6.51x more

AWS Trainium2

Compute (FP32-eq)

0.07x

Trainium2 is 14.72x faster

NVIDIA GeForce RTX 5070 Ti

Compute (FP32-eq)

0.53x

GeForce RTX 5070 Ti is 1.90x faster

FP32 Compute

0.53x

GeForce RTX 5070 Ti is 1.90x faster

VRAM

2.00x

Need 2.00x GeForce RTX 5070 Ti

Memory Bandwidth

1.37x

Need 1.37x GeForce RTX 5070 Ti

To match 1x NVIDIA B200

AMD MI100

Compute (FP32-eq)

48.70x

Need 48.70x MI100

FP32 Compute

3.90x

Need 3.90x MI100

VRAM

6.00x

Need 6.00x MI100

Memory Bandwidth

6.51x

Need 6.51x MI100

AWS Trainium2

Compute (FP32-eq)

3.31x

Need 3.31x Trainium2

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

To match 1x AWS Trainium2

AMD MI100

Compute (FP32-eq)

14.72x

Need 14.72x MI100

NVIDIA B200

Compute (FP32-eq)

0.30x

B200 is 3.31x faster

NVIDIA GeForce RTX 5070 Ti

Compute (FP32-eq)

7.73x

Need 7.73x GeForce RTX 5070 Ti

To match 1x NVIDIA GeForce RTX 5070 Ti

AMD MI100

Compute (FP32-eq)

1.90x

Need 1.90x MI100

FP32 Compute

1.90x

Need 1.90x MI100

VRAM

0.50x

MI100 has 2.00x more

Memory Bandwidth

0.73x

MI100 has 1.37x more

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

AWS Trainium2

Compute (FP32-eq)

0.13x

Trainium2 is 7.73x faster

Pricing

Price Type	MI100	B200	Trainium2	GeForce RTX 5070 Ti
CAPEX (Street Price)	—	$70,000	—	—
OPEX (per hour)	—	—	$1.34/hr	—
Price per TFLOPs (FP32-eq)	—	$62	—	—