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 MI250X	SambaNova SN40L	AMD MI300X	NVIDIA RTX 6000 Ada Generation
Architecture	CDNA 2	Reconfigurable Dataflow Unit (RDU)	CDNA 3	Ada Lovelace
Form Factor	OAM	—	OAM	PCIe
VRAM	128 GB	640 GB	192 GB	48 GB
Memory Bandwidth	3,277 GB/s	—	5,300 GB/s	960 GB/s
TFLOPs (FP32)	47.9	—	163	91.1
TFLOPs (FP16)	383	—	1,307	182.2
TFLOPs	383	—	1,307	182.2
TFLOPs (FP8)	—	—	2,614	—
TDP	560 W	700 W	750 W	300 W
Launch Date	Nov 2021	Jan 2024	Dec 2023	Dec 2022

Efficiency Metrics

Metric	MI250X	SN40L	MI300X	RTX 6000 Ada Generation
TFLOPs per Watt (FP32-eq)	0.34	—	0.87	0.30
Memory Bandwidth per GB	25.6 GB/s	—	27.6 GB/s	20.0 GB/s

Performance Equivalence

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

To match 1x AMD MI250X

SambaNova SN40L

VRAM

0.20x

SN40L has 5.00x more

AMD MI300X

Compute (FP32-eq)

0.29x

MI300X is 3.41x faster

FP32 Compute

0.29x

MI300X is 3.40x faster

VRAM

0.67x

MI300X has 1.50x more

Memory Bandwidth

0.62x

MI300X has 1.62x more

NVIDIA RTX 6000 Ada Generation

Compute (FP32-eq)

2.10x

Need 2.10x RTX 6000 Ada Generation

FP32 Compute

0.53x

RTX 6000 Ada Generation is 1.90x faster

VRAM

2.67x

Need 2.67x RTX 6000 Ada Generation

Memory Bandwidth

3.41x

Need 3.41x RTX 6000 Ada Generation

To match 1x SambaNova SN40L

AMD MI250X

VRAM

5.00x

Need 5.00x MI250X

AMD MI300X

VRAM

3.33x

Need 3.33x MI300X

NVIDIA RTX 6000 Ada Generation

VRAM

13.33x

Need 13.33x RTX 6000 Ada Generation

To match 1x AMD MI300X

AMD MI250X

Compute (FP32-eq)

3.41x

Need 3.41x MI250X

FP32 Compute

3.40x

Need 3.40x MI250X

VRAM

1.50x

Need 1.50x MI250X

Memory Bandwidth

1.62x

Need 1.62x MI250X

SambaNova SN40L

VRAM

0.30x

SN40L has 3.33x more

NVIDIA RTX 6000 Ada Generation

Compute (FP32-eq)

7.17x

Need 7.17x RTX 6000 Ada Generation

FP32 Compute

1.79x

Need 1.79x RTX 6000 Ada Generation

VRAM

4.00x

Need 4.00x RTX 6000 Ada Generation

Memory Bandwidth

5.52x

Need 5.52x RTX 6000 Ada Generation

To match 1x NVIDIA RTX 6000 Ada Generation

AMD MI250X

Compute (FP32-eq)

0.48x

MI250X is 2.10x faster

FP32 Compute

1.90x

Need 1.90x MI250X

VRAM

0.38x

MI250X has 2.67x more

Memory Bandwidth

0.29x

MI250X has 3.41x more

SambaNova SN40L

VRAM

0.07x

SN40L has 13.33x more

AMD MI300X

Compute (FP32-eq)

0.14x

MI300X is 7.17x faster

FP32 Compute

0.56x

MI300X is 1.79x faster

VRAM

0.25x

MI300X has 4.00x more

Memory Bandwidth

0.18x

MI300X has 5.52x more

Pricing

Price Type	MI250X	SN40L	MI300X	RTX 6000 Ada Generation
CAPEX (Street Price)	$12,000	—	$35,000	—
OPEX (per hour)	$2.00/hr	—	$10.40/hr	—
Price per TFLOPs (FP32-eq)	$63	—	$54	—