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 128 XPUs • 5 selected

Alibaba

Hanguang 800

AMD

MI100

23.1 TFLOPs

AMD

MI210

181 TFLOPs

AMD

MI250X

383 TFLOPs

AMD

MI300X

1,307 TFLOPs

AMD

MI325X

1,400 TFLOPs

AMD

MI350X

2,100 TFLOPs

AMD

MI355X

5,300 TFLOPs

AMD

Radeon Pro V520

23.04 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

FuriosaAI

RNGD (Renegade)

256 TFLOPs

FuriosaAI

Warboy

Google

TPU v4

275 TFLOPs

Google

TPU v5e

197 TFLOPs

Google

TPU v5p

459 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	Google TPU v5p	AMD MI325X	NVIDIA GeForce RTX 4070 Ti Super	NVIDIA A100 SXM	AMD MI355X
Architecture	TPU v5	CDNA 3.5	Ada Lovelace	Ampere	CDNA 4
Form Factor	Mezzanine	OAM	PCIe	SXM	OAM
VRAM	95 GB	256 GB	16 GB	80 GB	288 GB
Memory Bandwidth	—	6,000 GB/s	672 GB/s	2,039 GB/s	8,000 GB/s
TFLOPs (FP32)	—	207	—	19.5	320
TFLOPs (FP16)	—	1,400	—	312	—
TFLOPs	459	1,400	88	312	5,300
TFLOPs (FP8)	—	2,800	—	—	—
TDP	400 W	800 W	285 W	400 W	1000 W
Launch Date	Aug 2023	Oct 2024	Jan 2024	May 2020	Jun 2025

Efficiency Metrics

Metric	TPU v5p	MI325X	GeForce RTX 4070 Ti Super	A100 SXM	MI355X
TFLOPs per Watt (FP32-eq)	0.57	0.88	0.15	0.39	2.65
Memory Bandwidth per GB	—	23.4 GB/s	42.0 GB/s	25.5 GB/s	27.8 GB/s

Performance Equivalence

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

To match 1x Google TPU v5p

AMD MI325X

Compute (FP32-eq)

0.33x

MI325X is 3.05x faster

VRAM

0.37x

MI325X has 2.69x more

NVIDIA GeForce RTX 4070 Ti Super

Compute (FP32-eq)

5.22x

Need 5.22x GeForce RTX 4070 Ti Super

VRAM

5.94x

Need 5.94x GeForce RTX 4070 Ti Super

NVIDIA A100 SXM

Compute (FP32-eq)

1.47x

Need 1.47x A100 SXM

VRAM

1.19x

Need 1.19x A100 SXM

AMD MI355X

Compute (FP32-eq)

0.09x

MI355X is 11.55x faster

VRAM

0.33x

MI355X has 3.03x more

To match 1x AMD MI325X

Google TPU v5p

Compute (FP32-eq)

3.05x

Need 3.05x TPU v5p

VRAM

2.69x

Need 2.69x TPU v5p

NVIDIA GeForce RTX 4070 Ti Super

Compute (FP32-eq)

15.91x

Need 15.91x GeForce RTX 4070 Ti Super

VRAM

16.00x

Need 16.00x GeForce RTX 4070 Ti Super

Memory Bandwidth

8.93x

Need 8.93x GeForce RTX 4070 Ti Super

NVIDIA A100 SXM

Compute (FP32-eq)

4.49x

Need 4.49x A100 SXM

FP32 Compute

10.62x

Need 10.62x A100 SXM

VRAM

3.20x

Need 3.20x A100 SXM

Memory Bandwidth

2.94x

Need 2.94x A100 SXM

AMD MI355X

Compute (FP32-eq)

0.26x

MI355X is 3.79x faster

FP32 Compute

0.65x

MI355X is 1.55x faster

VRAM

0.89x

MI355X has 1.13x more

Memory Bandwidth

0.75x

MI355X has 1.33x more

To match 1x NVIDIA GeForce RTX 4070 Ti Super

Google TPU v5p

Compute (FP32-eq)

0.19x

TPU v5p is 5.22x faster

VRAM

0.17x

TPU v5p has 5.94x more

AMD MI325X

Compute (FP32-eq)

0.06x

MI325X is 15.91x faster

VRAM

0.06x

MI325X has 16.00x more

Memory Bandwidth

0.11x

MI325X has 8.93x more

NVIDIA A100 SXM

Compute (FP32-eq)

0.28x

A100 SXM is 3.55x faster

VRAM

0.20x

A100 SXM has 5.00x more

Memory Bandwidth

0.33x

A100 SXM has 3.03x more

AMD MI355X

Compute (FP32-eq)

0.02x

MI355X is 60.23x faster

VRAM

0.06x

MI355X has 18.00x more

Memory Bandwidth

0.08x

MI355X has 11.90x more

To match 1x NVIDIA A100 SXM

Google TPU v5p

Compute (FP32-eq)

0.68x

TPU v5p is 1.47x faster

VRAM

0.84x

TPU v5p has 1.19x more

AMD MI325X

Compute (FP32-eq)

0.22x

MI325X is 4.49x faster

FP32 Compute

0.09x

MI325X is 10.62x faster

VRAM

0.31x

MI325X has 3.20x more

Memory Bandwidth

0.34x

MI325X has 2.94x more

NVIDIA GeForce RTX 4070 Ti Super

Compute (FP32-eq)

3.55x

Need 3.55x GeForce RTX 4070 Ti Super

VRAM

5.00x

Need 5.00x GeForce RTX 4070 Ti Super

Memory Bandwidth

3.03x

Need 3.03x GeForce RTX 4070 Ti Super

AMD MI355X

Compute (FP32-eq)

0.06x

MI355X is 16.99x faster

FP32 Compute

0.06x

MI355X is 16.41x faster

VRAM

0.28x

MI355X has 3.60x more

Memory Bandwidth

0.25x

MI355X has 3.92x more

To match 1x AMD MI355X

Google TPU v5p

Compute (FP32-eq)

11.55x

Need 11.55x TPU v5p

VRAM

3.03x

Need 3.03x TPU v5p

AMD MI325X

Compute (FP32-eq)

3.79x

Need 3.79x MI325X

FP32 Compute

1.55x

Need 1.55x MI325X

VRAM

1.13x

Need 1.13x MI325X

Memory Bandwidth

1.33x

Need 1.33x MI325X

NVIDIA GeForce RTX 4070 Ti Super

Compute (FP32-eq)

60.23x

Need 60.23x GeForce RTX 4070 Ti Super

VRAM

18.00x

Need 18.00x GeForce RTX 4070 Ti Super

Memory Bandwidth

11.90x

Need 11.90x GeForce RTX 4070 Ti Super

NVIDIA A100 SXM

Compute (FP32-eq)

16.99x

Need 16.99x A100 SXM

FP32 Compute

16.41x

Need 16.41x A100 SXM

VRAM

3.60x

Need 3.60x A100 SXM

Memory Bandwidth

3.92x

Need 3.92x A100 SXM

Pricing

Price Type	TPU v5p	MI325X	GeForce RTX 4070 Ti Super	A100 SXM	MI355X
CAPEX (Street Price)	—	$45,000	—	$15,000	—
OPEX (per hour)	$5.00/hr	$36.92/hr	$0.12/hr	$4.05/hr	$20.72/hr
Price per TFLOPs (FP32-eq)	—	$64	—	$96	—