GeForce RTX 4090 and Tesla T40 24 GB - video card comparison

GeForce RTX 4090

Tesla T40 24 GB

Architecture

Ada Lovelace

Turing (2018−2021)

GPU code name

AD102-300-A1

TU102

Market segment

Desktop

Workstation

Release date

September 2022

Interface

PCIe 4.0 x16

PCIe 3.0 x16

Supplementary power connectors

1x 16-pin

1x 6-pin + 1x 8-pin

Boost clock speed

2520 MHz

1560 MHz

Core clock speed

2235 MHz

Manufacturing process technology

5 nm

12 nm

Number of transistors

76,300 million

18,600 million

Pipelines / CUDA cores

16384

4608

Power consumption (TDP)

450 Watt

260 Watt

Texture fill rate

1,290

449.3

Memory type

GDDR6X

GDDR6

Maximum RAM amount

24 GB

Memory bus width

384 Bit

Memory clock speed

21000 MHz

13 GB/s

Memory bandwidth

1,008 GB/s

624.0 GB/s

Shared memory

Display Connectors

1x HDMI 2.1, 3x DisplayPort 1.4a

No outputs

HDMI

DirectX

12 Ultimate (12_2)

CUDA

8.9

7.5

OpenCL

3.0

OpenGL

4.6

Shader Model

6.6

6.7

Vulkan

1.3