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 MI210	AMD MI350X	Alibaba Hanguang 800	NVIDIA GeForce RTX 4070 Super
Architecture	CDNA 2	CDNA 4	XuanTie	Ada Lovelace
Form Factor	PCIe	OAM	—	PCIe
VRAM	64 GB	288 GB	—	12 GB
Memory Bandwidth	1,638 GB/s	—	—	504 GB/s
TFLOPs (FP32)	45.3	—	—	35.5
TFLOPs (FP16)	181	—	—	71
TFLOPs	181	2,100	—	71
TFLOPs (FP8)	—	—	—	—
TDP	300 W	850 W	150 W	220 W
Launch Date	Jan 2022	Jun 2025	Sep 2019	Jan 2024

Efficiency Metrics

Metric	MI210	MI350X	Hanguang 800	GeForce RTX 4070 Super
TFLOPs per Watt (FP32-eq)	0.30	1.24	—	0.16
Memory Bandwidth per GB	25.6 GB/s	—	—	42.0 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 MI350X

Compute (FP32-eq)

0.09x

MI350X is 11.60x faster

VRAM

0.22x

MI350X has 4.50x more

NVIDIA GeForce RTX 4070 Super

Compute (FP32-eq)

2.55x

Need 2.55x GeForce RTX 4070 Super

FP32 Compute

1.28x

Need 1.28x GeForce RTX 4070 Super

VRAM

5.33x

Need 5.33x GeForce RTX 4070 Super

Memory Bandwidth

3.25x

Need 3.25x GeForce RTX 4070 Super

To match 1x AMD MI350X

AMD MI210

Compute (FP32-eq)

11.60x

Need 11.60x MI210

VRAM

4.50x

Need 4.50x MI210

NVIDIA GeForce RTX 4070 Super

Compute (FP32-eq)

29.58x

Need 29.58x GeForce RTX 4070 Super

VRAM

24.00x

Need 24.00x GeForce RTX 4070 Super

To match 1x NVIDIA GeForce RTX 4070 Super

AMD MI210

Compute (FP32-eq)

0.39x

MI210 is 2.55x faster

FP32 Compute

0.78x

MI210 is 1.28x faster

VRAM

0.19x

MI210 has 5.33x more

Memory Bandwidth

0.31x

MI210 has 3.25x more

AMD MI350X

Compute (FP32-eq)

0.03x

MI350X is 29.58x faster

VRAM

0.04x

MI350X has 24.00x more

Pricing

Price Type	MI210	MI350X	Hanguang 800	GeForce RTX 4070 Super
CAPEX (Street Price)	$6,000	—	—	—
OPEX (per hour)	—	—	—	—
Price per TFLOPs (FP32-eq)	$66	—	—	—