CMP 40HX and L40 - video card comparison

CMP 40HX

L40

Architecture

Turing (2018−2021)

Ada Lovelace

GPU code name

TU106

AD102

Market segment

Desktop

Desktop

Release date

February 2021

October 2022

Interface

PCIe 3.0 x16

PCIe 4.0 x16

Supplementary power connectors

1x 8-pin

1x 16-pin

Boost clock speed

1650 MHz

2490 MHz

Manufacturing process technology

12 nm

4 nm

Number of transistors

10,800 million

76,300 million

Pipelines / CUDA cores

2304

18176

Power consumption (TDP)

185 Watt

300 Watt

Texture fill rate

237.6

1,414

Memory type

GDDR6

GDDR6

Maximum RAM amount

8 GB

48 GB

Memory bus width

256 Bit

384 Bit

Memory clock speed

14 GB/s

18 GB/s

Memory bandwidth

448.0 GB/s

864.0 GB/s

Display Connectors

No outputs

4x DisplayPort 1.4a

DirectX

12 Ultimate (12_2)

12 Ultimate (12_2)

CUDA

7.5

8.9

OpenCL

3.0

3.0

OpenGL

4.6

4.6

Shader Model

6.6

6.7

Vulkan

1.2

1.3