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 A100 SXM	NVIDIA RTX PRO 6000 Blackwell Max-Q	AMD MI300A
Architecture	Ampere	Blackwell	CDNA 3
Form Factor	SXM	PCIe	OAM
VRAM	80 GB	96 GB	128 GB
Memory Bandwidth	2,039 GB/s	1,792 GB/s	5,300 GB/s
TFLOPs (FP32)	19.5	62.5	122.6
TFLOPs (FP16)	312	125	980.6
TFLOPs	312	125	980.6
TFLOPs (FP8)	—	—	—
TDP	400 W	300 W	760 W
Launch Date	May 2020	Mar 2025	Dec 2023

Efficiency Metrics

Metric	A100 SXM	RTX PRO 6000 Blackwell Max-Q	MI300A
TFLOPs per Watt (FP32-eq)	0.39	0.21	0.65
Memory Bandwidth per GB	25.5 GB/s	18.7 GB/s	41.4 GB/s

Performance Equivalence

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

To match 1x NVIDIA A100 SXM

NVIDIA RTX PRO 6000 Blackwell Max-Q

Compute (FP32-eq)

2.50x

Need 2.50x RTX PRO 6000 Blackwell Max-Q

FP32 Compute

0.31x

RTX PRO 6000 Blackwell Max-Q is 3.21x faster

VRAM

0.83x

RTX PRO 6000 Blackwell Max-Q has 1.20x more

Memory Bandwidth

1.14x

Need 1.14x RTX PRO 6000 Blackwell Max-Q

AMD MI300A

Compute (FP32-eq)

0.32x

MI300A is 3.14x faster

FP32 Compute

0.16x

MI300A is 6.29x faster

VRAM

0.63x

MI300A has 1.60x more

Memory Bandwidth

0.38x

MI300A has 2.60x more

To match 1x NVIDIA RTX PRO 6000 Blackwell Max-Q

NVIDIA A100 SXM

Compute (FP32-eq)

0.40x

A100 SXM is 2.50x faster

FP32 Compute

3.21x

Need 3.21x A100 SXM

VRAM

1.20x

Need 1.20x A100 SXM

Memory Bandwidth

0.88x

A100 SXM has 1.14x more

AMD MI300A

Compute (FP32-eq)

0.13x

MI300A is 7.84x faster

FP32 Compute

0.51x

MI300A is 1.96x faster

VRAM

0.75x

MI300A has 1.33x more

Memory Bandwidth

0.34x

MI300A has 2.96x more

To match 1x AMD MI300A

NVIDIA A100 SXM

Compute (FP32-eq)

3.14x

Need 3.14x A100 SXM

FP32 Compute

6.29x

Need 6.29x A100 SXM

VRAM

1.60x

Need 1.60x A100 SXM

Memory Bandwidth

2.60x

Need 2.60x A100 SXM

NVIDIA RTX PRO 6000 Blackwell Max-Q

Compute (FP32-eq)

7.84x

Need 7.84x RTX PRO 6000 Blackwell Max-Q

FP32 Compute

1.96x

Need 1.96x RTX PRO 6000 Blackwell Max-Q

VRAM

1.33x

Need 1.33x RTX PRO 6000 Blackwell Max-Q

Memory Bandwidth

2.96x

Need 2.96x RTX PRO 6000 Blackwell Max-Q

Pricing

Price Type	A100 SXM	RTX PRO 6000 Blackwell Max-Q	MI300A
CAPEX (Street Price)	$15,000	—	—
OPEX (per hour)	$4.05/hr	—	—
Price per TFLOPs (FP32-eq)	$96	—	—