Quadro RTX 8000 and Tesla K40t - video card comparison

Quadro RTX 8000

Tesla K40t

Architecture

Turing (2018−2021)

Kepler (2012−2018)

GPU code name

TU102

GK110B

Market segment

Workstation

Workstation

Release date

August 2018

November 2013

Interface

PCIe 3.0 x16

PCIe 3.0 x16

Supplementary power connectors

1x 6-pin + 1x 8-pin

-

Boost clock speed

1770 MHz

876 MHz

Core clock speed

1395 MHz

745 MHz

Manufacturing process technology

12 nm

28 nm

Number of transistors

18,600 million

7,080 million

Pipelines / CUDA cores

4608

2880

Power consumption (TDP)

260 Watt

245 Watt

Texture fill rate

509.8

210.2

Memory type

GDDR6

GDDR5

Maximum RAM amount

48 GB

12 GB

Memory bus width

384 Bit

384 Bit

Memory clock speed

14000 MHz

6008 MHz

Memory bandwidth

672.0 GB/s

288.4 GB/s

Display Connectors

4x DisplayPort, 1x USB Type-C

No outputs

DirectX

12 Ultimate (12_1)

12 (11_1)

CUDA

7.5

3.5

OpenCL

2.0

1.2

OpenGL

4.6

4.6

Shader Model

6.5

5.1

Vulkan

1.2.131

1.1.126