Compare XPUs

Select up to 5 XPUs to compare side-by-side

Select XPUs to Compare

Clear all (6)Jump to results

Filter by Vendor

Showing 72 XPUs • 6 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 L4	NVIDIA A40	NVIDIA A10	AMD MI250X	AMD MI350X	FuriosaAI Warboy
Architecture	Ada Lovelace	Ampere	Ampere	CDNA 2	CDNA 4	FuriosaAI NPU
Form Factor	PCIe	PCIe	PCIe	OAM	OAM	—
VRAM	24 GB	48 GB	24 GB	128 GB	288 GB	32 GB
Memory Bandwidth	300 GB/s	696 GB/s	600 GB/s	3,277 GB/s	—	—
TFLOPs (FP32)	30.3	37.4	31.2	47.9	—	—
TFLOPs (FP16)	242	150	125	383	—	—
TFLOPs	121	150	125	383	2,100	—
TFLOPs (FP8)	485	—	—	—	—	—
TDP	72 W	300 W	150 W	560 W	850 W	80 W
Launch Date	Mar 2023	Oct 2020	Apr 2021	Nov 2021	Jun 2025	Oct 2021

Efficiency Metrics

Metric	L4	A40	A10	MI250X	MI350X	Warboy
TFLOPs per Watt (FP32-eq)	0.84	0.25	0.42	0.34	1.24	—
Memory Bandwidth per GB	12.5 GB/s	14.5 GB/s	25.0 GB/s	25.6 GB/s	—	—

Performance Equivalence

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

To match 1x NVIDIA L4

NVIDIA A40

Compute (FP32-eq)

0.81x

A40 is 1.24x faster

FP32 Compute

0.81x

A40 is 1.23x faster

VRAM

0.50x

A40 has 2.00x more

Memory Bandwidth

0.43x

A40 has 2.32x more

NVIDIA A10

Compute (FP32-eq)

0.97x

A10 is 1.03x faster

FP32 Compute

0.97x

A10 is 1.03x faster

VRAM

1.00x

A10 has 1.00x more

Memory Bandwidth

0.50x

A10 has 2.00x more

AMD MI250X

Compute (FP32-eq)

0.32x

MI250X is 3.17x faster

FP32 Compute

0.63x

MI250X is 1.58x faster

VRAM

0.19x

MI250X has 5.33x more

Memory Bandwidth

0.09x

MI250X has 10.92x more

AMD MI350X

Compute (FP32-eq)

0.06x

MI350X is 17.36x faster

VRAM

0.08x

MI350X has 12.00x more

FuriosaAI Warboy

VRAM

0.75x

Warboy has 1.33x more

To match 1x NVIDIA A40

NVIDIA L4

Compute (FP32-eq)

1.24x

Need 1.24x L4

FP32 Compute

1.23x

Need 1.23x L4

VRAM

2.00x

Need 2.00x L4

Memory Bandwidth

2.32x

Need 2.32x L4

NVIDIA A10

Compute (FP32-eq)

1.20x

Need 1.20x A10

FP32 Compute

1.20x

Need 1.20x A10

VRAM

2.00x

Need 2.00x A10

Memory Bandwidth

1.16x

Need 1.16x A10

AMD MI250X

Compute (FP32-eq)

0.39x

MI250X is 2.55x faster

FP32 Compute

0.78x

MI250X is 1.28x faster

VRAM

0.38x

MI250X has 2.67x more

Memory Bandwidth

0.21x

MI250X has 4.71x more

AMD MI350X

Compute (FP32-eq)

0.07x

MI350X is 14.00x faster

VRAM

0.17x

MI350X has 6.00x more

FuriosaAI Warboy

VRAM

1.50x

Need 1.50x Warboy

To match 1x NVIDIA A10

NVIDIA L4

Compute (FP32-eq)

1.03x

Need 1.03x L4

FP32 Compute

1.03x

Need 1.03x L4

VRAM

1.00x

L4 has 1.00x more

Memory Bandwidth

2.00x

Need 2.00x L4

NVIDIA A40

Compute (FP32-eq)

0.83x

A40 is 1.20x faster

FP32 Compute

0.83x

A40 is 1.20x faster

VRAM

0.50x

A40 has 2.00x more

Memory Bandwidth

0.86x

A40 has 1.16x more

AMD MI250X

Compute (FP32-eq)

0.33x

MI250X is 3.06x faster

FP32 Compute

0.65x

MI250X is 1.54x faster

VRAM

0.19x

MI250X has 5.33x more

Memory Bandwidth

0.18x

MI250X has 5.46x more

AMD MI350X

Compute (FP32-eq)

0.06x

MI350X is 16.80x faster

VRAM

0.08x

MI350X has 12.00x more

FuriosaAI Warboy

VRAM

0.75x

Warboy has 1.33x more

To match 1x AMD MI250X

NVIDIA L4

Compute (FP32-eq)

3.17x

Need 3.17x L4

FP32 Compute

1.58x

Need 1.58x L4

VRAM

5.33x

Need 5.33x L4

Memory Bandwidth

10.92x

Need 10.92x L4

NVIDIA A40

Compute (FP32-eq)

2.55x

Need 2.55x A40

FP32 Compute

1.28x

Need 1.28x A40

VRAM

2.67x

Need 2.67x A40

Memory Bandwidth

4.71x

Need 4.71x A40

NVIDIA A10

Compute (FP32-eq)

3.06x

Need 3.06x A10

FP32 Compute

1.54x

Need 1.54x A10

VRAM

5.33x

Need 5.33x A10

Memory Bandwidth

5.46x

Need 5.46x A10

AMD MI350X

Compute (FP32-eq)

0.18x

MI350X is 5.48x faster

VRAM

0.44x

MI350X has 2.25x more

FuriosaAI Warboy

VRAM

4.00x

Need 4.00x Warboy

To match 1x AMD MI350X

NVIDIA L4

Compute (FP32-eq)

17.36x

Need 17.36x L4

VRAM

12.00x

Need 12.00x L4

NVIDIA A40

Compute (FP32-eq)

14.00x

Need 14.00x A40

VRAM

6.00x

Need 6.00x A40

NVIDIA A10

Compute (FP32-eq)

16.80x

Need 16.80x A10

VRAM

12.00x

Need 12.00x A10

AMD MI250X

Compute (FP32-eq)

5.48x

Need 5.48x MI250X

VRAM

2.25x

Need 2.25x MI250X

FuriosaAI Warboy

VRAM

9.00x

Need 9.00x Warboy

To match 1x FuriosaAI Warboy

NVIDIA L4

VRAM

1.33x

Need 1.33x L4

NVIDIA A40

VRAM

0.67x

A40 has 1.50x more

NVIDIA A10

VRAM

1.33x

Need 1.33x A10

AMD MI250X

VRAM

0.25x

MI250X has 4.00x more

AMD MI350X

VRAM

0.11x

MI350X has 9.00x more

Pricing

Price Type	L4	A40	A10	MI250X	MI350X	Warboy
CAPEX (Street Price)	$4,000	—	—	$12,000	—	—
OPEX (per hour)	$0.80/hr	—	—	$2.00/hr	—	—
Price per TFLOPs (FP32-eq)	$66	—	—	$63	—	—