Compare XPUs

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

Select XPUs to Compare

Clear all (3)Jump to results

Filter by Vendor

Showing 72 XPUs • 3 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 A10	AMD MI300X	NVIDIA GeForce RTX 4070 Super
Architecture	Ampere	CDNA 3	Ada Lovelace
Form Factor	PCIe	OAM	PCIe
VRAM	24 GB	192 GB	12 GB
Memory Bandwidth	600 GB/s	5,300 GB/s	504 GB/s
TFLOPs (FP32)	31.2	163	35.5
TFLOPs (FP16)	125	1,307	71
TFLOPs	125	1,307	71
TFLOPs (FP8)	—	2,614	—
TDP	150 W	750 W	220 W
Launch Date	Apr 2021	Dec 2023	Jan 2024

Efficiency Metrics

Metric	A10	MI300X	GeForce RTX 4070 Super
TFLOPs per Watt (FP32-eq)	0.42	0.87	0.16
Memory Bandwidth per GB	25.0 GB/s	27.6 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 NVIDIA A10

AMD MI300X

Compute (FP32-eq)

0.10x

MI300X is 10.46x faster

FP32 Compute

0.19x

MI300X is 5.22x faster

VRAM

0.13x

MI300X has 8.00x more

Memory Bandwidth

0.11x

MI300X has 8.83x more

NVIDIA GeForce RTX 4070 Super

Compute (FP32-eq)

1.76x

Need 1.76x GeForce RTX 4070 Super

FP32 Compute

0.88x

GeForce RTX 4070 Super is 1.14x faster

VRAM

2.00x

Need 2.00x GeForce RTX 4070 Super

Memory Bandwidth

1.19x

Need 1.19x GeForce RTX 4070 Super

To match 1x AMD MI300X

NVIDIA A10

Compute (FP32-eq)

10.46x

Need 10.46x A10

FP32 Compute

5.22x

Need 5.22x A10

VRAM

8.00x

Need 8.00x A10

Memory Bandwidth

8.83x

Need 8.83x A10

NVIDIA GeForce RTX 4070 Super

Compute (FP32-eq)

18.41x

Need 18.41x GeForce RTX 4070 Super

FP32 Compute

4.59x

Need 4.59x GeForce RTX 4070 Super

VRAM

16.00x

Need 16.00x GeForce RTX 4070 Super

Memory Bandwidth

10.52x

Need 10.52x GeForce RTX 4070 Super

To match 1x NVIDIA GeForce RTX 4070 Super

NVIDIA A10

Compute (FP32-eq)

0.57x

A10 is 1.76x faster

FP32 Compute

1.14x

Need 1.14x A10

VRAM

0.50x

A10 has 2.00x more

Memory Bandwidth

0.84x

A10 has 1.19x more

AMD MI300X

Compute (FP32-eq)

0.05x

MI300X is 18.41x faster

FP32 Compute

0.22x

MI300X is 4.59x faster

VRAM

0.06x

MI300X has 16.00x more

Memory Bandwidth

0.10x

MI300X has 10.52x more

Pricing

Price Type	A10	MI300X	GeForce RTX 4070 Super
CAPEX (Street Price)	—	$35,000	—
OPEX (per hour)	—	$10.40/hr	—
Price per TFLOPs (FP32-eq)	—	$54	—