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	Meta MTIA v1	NVIDIA GB200 Superchip	NVIDIA GB200 NVL72	AMD MI300A
Architecture	Meta Training & Inference Accelerator	Blackwell	Blackwell	CDNA 3
Form Factor	—	SXM	Custom	OAM
VRAM	128 GB	384 GB	13824 GB	128 GB
Memory Bandwidth	—	8,000 GB/s	130,000 GB/s	5,300 GB/s
TFLOPs (FP32)	—	90	6,480	122.6
TFLOPs (FP16)	—	5,000	360,000	980.6
TFLOPs	—	5,000	360,000	980.6
TFLOPs (FP8)	—	—	—	—
TDP	400 W	2700 W	120000 W	760 W
Launch Date	May 2023	Mar 2024	Mar 2024	Dec 2023

Efficiency Metrics

Metric	MTIA v1	GB200 Superchip	GB200 NVL72	MI300A
TFLOPs per Watt (FP32-eq)	—	0.93	1.50	0.65
Memory Bandwidth per GB	—	20.8 GB/s	9.4 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 Meta MTIA v1

NVIDIA GB200 Superchip

VRAM

0.33x

GB200 Superchip has 3.00x more

NVIDIA GB200 NVL72

VRAM

0.01x

GB200 NVL72 has 108.00x more

AMD MI300A

VRAM

1.00x

MI300A has 1.00x more

To match 1x NVIDIA GB200 Superchip

Meta MTIA v1

VRAM

3.00x

Need 3.00x MTIA v1

NVIDIA GB200 NVL72

Compute (FP32-eq)

0.01x

GB200 NVL72 is 72.00x faster

FP32 Compute

0.01x

GB200 NVL72 is 72.00x faster

VRAM

0.03x

GB200 NVL72 has 36.00x more

Memory Bandwidth

0.06x

GB200 NVL72 has 16.25x more

AMD MI300A

Compute (FP32-eq)

5.10x

Need 5.10x MI300A

FP32 Compute

0.73x

MI300A is 1.36x faster

VRAM

3.00x

Need 3.00x MI300A

Memory Bandwidth

1.51x

Need 1.51x MI300A

To match 1x NVIDIA GB200 NVL72

Meta MTIA v1

VRAM

108.00x

Need 108.00x MTIA v1

NVIDIA GB200 Superchip

Compute (FP32-eq)

72.00x

Need 72.00x GB200 Superchip

FP32 Compute

72.00x

Need 72.00x GB200 Superchip

VRAM

36.00x

Need 36.00x GB200 Superchip

Memory Bandwidth

16.25x

Need 16.25x GB200 Superchip

AMD MI300A

Compute (FP32-eq)

367.12x

Need 367.12x MI300A

FP32 Compute

52.85x

Need 52.85x MI300A

VRAM

108.00x

Need 108.00x MI300A

Memory Bandwidth

24.53x

Need 24.53x MI300A

To match 1x AMD MI300A

Meta MTIA v1

VRAM

1.00x

MTIA v1 has 1.00x more

NVIDIA GB200 Superchip

Compute (FP32-eq)

0.20x

GB200 Superchip is 5.10x faster

FP32 Compute

1.36x

Need 1.36x GB200 Superchip

VRAM

0.33x

GB200 Superchip has 3.00x more

Memory Bandwidth

0.66x

GB200 Superchip has 1.51x more

NVIDIA GB200 NVL72

Compute (FP32-eq)

0.00x

GB200 NVL72 is 367.12x faster

FP32 Compute

0.02x

GB200 NVL72 is 52.85x faster

VRAM

0.01x

GB200 NVL72 has 108.00x more

Memory Bandwidth

0.04x

GB200 NVL72 has 24.53x more

Pricing

Price Type	MTIA v1	GB200 Superchip	GB200 NVL72	MI300A
CAPEX (Street Price)	—	—	—	—
OPEX (per hour)	—	—	—	—
Price per TFLOPs (FP32-eq)	—	—	—	—