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 72 XPUs • 5 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	AMD MI210	Meta MTIA v1	Rebellions ATOM	NVIDIA H200 PCIe
Architecture	CDNA 2	CDNA 2	Meta Training & Inference Accelerator	Rebellions NPU	Hopper
Form Factor	OAM	PCIe	—	—	PCIe
VRAM	128 GB	64 GB	128 GB	—	141 GB
Memory Bandwidth	3,277 GB/s	1,638 GB/s	—	—	4,800 GB/s
TFLOPs (FP32)	47.9	45.3	—	—	51
TFLOPs (FP16)	383	181	—	—	1,513
TFLOPs	383	181	—	—	1,513
TFLOPs (FP8)	—	—	—	—	—
TDP	560 W	300 W	400 W	200 W	350 W
Launch Date	Nov 2021	Jan 2022	May 2023	Jun 2023	Mar 2024

Efficiency Metrics

Metric	MI250X	MI210	MTIA v1	ATOM	H200 PCIe
TFLOPs per Watt (FP32-eq)	0.34	0.30	—	—	2.16
Memory Bandwidth per GB	25.6 GB/s	25.6 GB/s	—	—	34.0 GB/s

Performance Equivalence

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

To match 1x AMD MI250X

AMD MI210

Compute (FP32-eq)

2.12x

Need 2.12x MI210

FP32 Compute

1.06x

Need 1.06x MI210

VRAM

2.00x

Need 2.00x MI210

Memory Bandwidth

2.00x

Need 2.00x MI210

Meta MTIA v1

VRAM

1.00x

MTIA v1 has 1.00x more

NVIDIA H200 PCIe

Compute (FP32-eq)

0.25x

H200 PCIe is 3.95x faster

FP32 Compute

0.94x

H200 PCIe is 1.06x faster

VRAM

0.91x

H200 PCIe has 1.10x more

Memory Bandwidth

0.68x

H200 PCIe has 1.46x more

To match 1x AMD MI210

AMD MI250X

Compute (FP32-eq)

0.47x

MI250X is 2.12x faster

FP32 Compute

0.95x

MI250X is 1.06x faster

VRAM

0.50x

MI250X has 2.00x more

Memory Bandwidth

0.50x

MI250X has 2.00x more

Meta MTIA v1

VRAM

0.50x

MTIA v1 has 2.00x more

NVIDIA H200 PCIe

Compute (FP32-eq)

0.12x

H200 PCIe is 8.36x faster

FP32 Compute

0.89x

H200 PCIe is 1.13x faster

VRAM

0.45x

H200 PCIe has 2.20x more

Memory Bandwidth

0.34x

H200 PCIe has 2.93x more

To match 1x Meta MTIA v1

AMD MI250X

VRAM

1.00x

MI250X has 1.00x more

AMD MI210

VRAM

2.00x

Need 2.00x MI210

NVIDIA H200 PCIe

VRAM

0.91x

H200 PCIe has 1.10x more

To match 1x NVIDIA H200 PCIe

AMD MI250X

Compute (FP32-eq)

3.95x

Need 3.95x MI250X

FP32 Compute

1.06x

Need 1.06x MI250X

VRAM

1.10x

Need 1.10x MI250X

Memory Bandwidth

1.46x

Need 1.46x MI250X

AMD MI210

Compute (FP32-eq)

8.36x

Need 8.36x MI210

FP32 Compute

1.13x

Need 1.13x MI210

VRAM

2.20x

Need 2.20x MI210

Memory Bandwidth

2.93x

Need 2.93x MI210

Meta MTIA v1

VRAM

1.10x

Need 1.10x MTIA v1

Pricing

Price Type	MI250X	MI210	MTIA v1	ATOM	H200 PCIe
CAPEX (Street Price)	$12,000	$6,000	—	—	—
OPEX (per hour)	$2.00/hr	—	—	—	—
Price per TFLOPs (FP32-eq)	$63	$66	—	—	—