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 128 XPUs • 5 selected

Alibaba

Hanguang 800

AMD

MI100

23.1 TFLOPs

AMD

MI210

181 TFLOPs

AMD

MI250X

383 TFLOPs

AMD

MI300X

1,307 TFLOPs

AMD

MI325X

1,400 TFLOPs

AMD

MI350X

2,100 TFLOPs

AMD

MI355X

5,300 TFLOPs

AMD

Radeon Pro V520

23.04 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

FuriosaAI

RNGD (Renegade)

256 TFLOPs

FuriosaAI

Warboy

Google

TPU v4

275 TFLOPs

Google

TPU v5e

197 TFLOPs

Google

TPU v5p

459 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	Tenstorrent Wormhole	Intel Data Center GPU Max 1100	NVIDIA GeForce RTX 5060 Ti	NVIDIA L40	NVIDIA T4
Architecture	Tensix Core	Ponte Vecchio	Blackwell	Ada Lovelace	Turing
Form Factor	—	OAM	PCIe	PCIe	PCIe
VRAM	24 GB	48 GB	16 GB	48 GB	16 GB
Memory Bandwidth	—	1,229 GB/s	544 GB/s	864 GB/s	320 GB/s
TFLOPs (FP32)	—	22	25	45	8.1
TFLOPs (FP16)	—	177	—	—	—
TFLOPs	364	177	88	181.05	65
TFLOPs (FP8)	—	—	—	—	—
TDP	160 W	300 W	220 W	300 W	70 W
Launch Date	Oct 2023	Jan 2023	Mar 2025	Oct 2022	Sep 2018

Efficiency Metrics

Metric	Wormhole	Data Center GPU Max 1100	GeForce RTX 5060 Ti	L40	T4
TFLOPs per Watt (FP32-eq)	1.14	0.29	0.20	0.30	0.46
Memory Bandwidth per GB	—	25.6 GB/s	34.0 GB/s	18.0 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 Tenstorrent Wormhole

Intel Data Center GPU Max 1100

Compute (FP32-eq)

2.06x

Need 2.06x Data Center GPU Max 1100

VRAM

0.50x

Data Center GPU Max 1100 has 2.00x more

NVIDIA GeForce RTX 5060 Ti

Compute (FP32-eq)

4.14x

Need 4.14x GeForce RTX 5060 Ti

VRAM

1.50x

Need 1.50x GeForce RTX 5060 Ti

NVIDIA L40

Compute (FP32-eq)

2.01x

Need 2.01x L40

VRAM

0.50x

L40 has 2.00x more

NVIDIA T4

Compute (FP32-eq)

5.60x

Need 5.60x T4

VRAM

1.50x

Need 1.50x T4

To match 1x Intel Data Center GPU Max 1100

Tenstorrent Wormhole

Compute (FP32-eq)

0.49x

Wormhole is 2.06x faster

VRAM

2.00x

Need 2.00x Wormhole

NVIDIA GeForce RTX 5060 Ti

Compute (FP32-eq)

2.01x

Need 2.01x GeForce RTX 5060 Ti

FP32 Compute

0.88x

GeForce RTX 5060 Ti is 1.14x faster

VRAM

3.00x

Need 3.00x GeForce RTX 5060 Ti

Memory Bandwidth

2.26x

Need 2.26x GeForce RTX 5060 Ti

NVIDIA L40

Compute (FP32-eq)

0.98x

L40 is 1.02x faster

FP32 Compute

0.49x

L40 is 2.05x faster

VRAM

1.00x

L40 has 1.00x more

Memory Bandwidth

1.42x

Need 1.42x L40

NVIDIA T4

Compute (FP32-eq)

2.72x

Need 2.72x T4

FP32 Compute

2.72x

Need 2.72x T4

VRAM

3.00x

Need 3.00x T4

Memory Bandwidth

3.84x

Need 3.84x T4

To match 1x NVIDIA GeForce RTX 5060 Ti

Tenstorrent Wormhole

Compute (FP32-eq)

0.24x

Wormhole is 4.14x faster

VRAM

0.67x

Wormhole has 1.50x more

Intel Data Center GPU Max 1100

Compute (FP32-eq)

0.50x

Data Center GPU Max 1100 is 2.01x faster

FP32 Compute

1.14x

Need 1.14x Data Center GPU Max 1100

VRAM

0.33x

Data Center GPU Max 1100 has 3.00x more

Memory Bandwidth

0.44x

Data Center GPU Max 1100 has 2.26x more

NVIDIA L40

Compute (FP32-eq)

0.49x

L40 is 2.06x faster

FP32 Compute

0.56x

L40 is 1.80x faster

VRAM

0.33x

L40 has 3.00x more

Memory Bandwidth

0.63x

L40 has 1.59x more

NVIDIA T4

Compute (FP32-eq)

1.35x

Need 1.35x T4

FP32 Compute

3.09x

Need 3.09x T4

VRAM

1.00x

T4 has 1.00x more

Memory Bandwidth

1.70x

Need 1.70x T4

To match 1x NVIDIA L40

Tenstorrent Wormhole

Compute (FP32-eq)

0.50x

Wormhole is 2.01x faster

VRAM

2.00x

Need 2.00x Wormhole

Intel Data Center GPU Max 1100

Compute (FP32-eq)

1.02x

Need 1.02x Data Center GPU Max 1100

FP32 Compute

2.05x

Need 2.05x Data Center GPU Max 1100

VRAM

1.00x

Data Center GPU Max 1100 has 1.00x more

Memory Bandwidth

0.70x

Data Center GPU Max 1100 has 1.42x more

NVIDIA GeForce RTX 5060 Ti

Compute (FP32-eq)

2.06x

Need 2.06x GeForce RTX 5060 Ti

FP32 Compute

1.80x

Need 1.80x GeForce RTX 5060 Ti

VRAM

3.00x

Need 3.00x GeForce RTX 5060 Ti

Memory Bandwidth

1.59x

Need 1.59x GeForce RTX 5060 Ti

NVIDIA T4

Compute (FP32-eq)

2.79x

Need 2.79x T4

FP32 Compute

5.56x

Need 5.56x T4

VRAM

3.00x

Need 3.00x T4

Memory Bandwidth

2.70x

Need 2.70x T4

To match 1x NVIDIA T4

Tenstorrent Wormhole

Compute (FP32-eq)

0.18x

Wormhole is 5.60x faster

VRAM

0.67x

Wormhole has 1.50x more

Intel Data Center GPU Max 1100

Compute (FP32-eq)

0.37x

Data Center GPU Max 1100 is 2.72x faster

FP32 Compute

0.37x

Data Center GPU Max 1100 is 2.72x faster

VRAM

0.33x

Data Center GPU Max 1100 has 3.00x more

Memory Bandwidth

0.26x

Data Center GPU Max 1100 has 3.84x more

NVIDIA GeForce RTX 5060 Ti

Compute (FP32-eq)

0.74x

GeForce RTX 5060 Ti is 1.35x faster

FP32 Compute

0.32x

GeForce RTX 5060 Ti is 3.09x faster

VRAM

1.00x

GeForce RTX 5060 Ti has 1.00x more

Memory Bandwidth

0.59x

GeForce RTX 5060 Ti has 1.70x more

NVIDIA L40

Compute (FP32-eq)

0.36x

L40 is 2.79x faster

FP32 Compute

0.18x

L40 is 5.56x faster

VRAM

0.33x

L40 has 3.00x more

Memory Bandwidth

0.37x

L40 has 2.70x more

Pricing

Price Type	Wormhole	Data Center GPU Max 1100	GeForce RTX 5060 Ti	L40	T4
CAPEX (Street Price)	—	$5,000	—	—	—
OPEX (per hour)	—	—	$0.09/hr	$0.69/hr	$0.27/hr
Price per TFLOPs (FP32-eq)	—	$56	—	—	—