Compare XPUs

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

Select XPUs to Compare

Clear all (5)Jump to results

Filter by Vendor

Showing 72 XPUs • 5 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 MI210	AMD MI100	Tenstorrent Grayskull	Alibaba Hanguang 800	AWS Trainium
Architecture	CDNA 2	CDNA	Tensix Core	XuanTie	Inferentia/Trainium
Form Factor	PCIe	PCIe	—	—	—
VRAM	64 GB	32 GB	8 GB	—	32 GB
Memory Bandwidth	1,638 GB/s	1,228 GB/s	—	—	—
TFLOPs (FP32)	45.3	23.1	—	—	—
TFLOPs (FP16)	181	184.6	—	—	—
TFLOPs	181	23.1	200	—	190
TFLOPs (FP8)	—	—	—	—	—
TDP	300 W	300 W	65 W	150 W	200 W
Launch Date	Jan 2022	Nov 2020	Mar 2021	Sep 2019	Nov 2021

Efficiency Metrics

Metric	MI210	MI100	Grayskull	Hanguang 800	Trainium
TFLOPs per Watt (FP32-eq)	0.30	0.08	1.54	—	0.47
Memory Bandwidth per GB	25.6 GB/s	38.4 GB/s	—	—	—

Performance Equivalence

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

To match 1x AMD MI210

AMD MI100

Compute (FP32-eq)

3.92x

Need 3.92x MI100

FP32 Compute

1.96x

Need 1.96x MI100

VRAM

2.00x

Need 2.00x MI100

Memory Bandwidth

1.33x

Need 1.33x MI100

Tenstorrent Grayskull

Compute (FP32-eq)

0.91x

Grayskull is 1.10x faster

VRAM

8.00x

Need 8.00x Grayskull

AWS Trainium

Compute (FP32-eq)

0.95x

Trainium is 1.05x faster

VRAM

2.00x

Need 2.00x Trainium

To match 1x AMD MI100

AMD MI210

Compute (FP32-eq)

0.26x

MI210 is 3.92x faster

FP32 Compute

0.51x

MI210 is 1.96x faster

VRAM

0.50x

MI210 has 2.00x more

Memory Bandwidth

0.75x

MI210 has 1.33x more

Tenstorrent Grayskull

Compute (FP32-eq)

0.23x

Grayskull is 4.33x faster

VRAM

4.00x

Need 4.00x Grayskull

AWS Trainium

Compute (FP32-eq)

0.24x

Trainium is 4.11x faster

VRAM

1.00x

Trainium has 1.00x more

To match 1x Tenstorrent Grayskull

AMD MI210

Compute (FP32-eq)

1.10x

Need 1.10x MI210

VRAM

0.13x

MI210 has 8.00x more

AMD MI100

Compute (FP32-eq)

4.33x

Need 4.33x MI100

VRAM

0.25x

MI100 has 4.00x more

AWS Trainium

Compute (FP32-eq)

1.05x

Need 1.05x Trainium

VRAM

0.25x

Trainium has 4.00x more

To match 1x AWS Trainium

AMD MI210

Compute (FP32-eq)

1.05x

Need 1.05x MI210

VRAM

0.50x

MI210 has 2.00x more

AMD MI100

Compute (FP32-eq)

4.11x

Need 4.11x MI100

VRAM

1.00x

MI100 has 1.00x more

Tenstorrent Grayskull

Compute (FP32-eq)

0.95x

Grayskull is 1.05x faster

VRAM

4.00x

Need 4.00x Grayskull

Pricing

Price Type	MI210	MI100	Grayskull	Hanguang 800	Trainium
CAPEX (Street Price)	$6,000	—	—	—	—
OPEX (per hour)	—	—	—	—	—
Price per TFLOPs (FP32-eq)	$66	—	—	—	—