GeForce RTX 3080 and Radeon RX 580 OEM - video card comparison

GeForce RTX 3080

Radeon RX 580 OEM

Architecture

Ampere (2020−2022)

GCN 4.0 (2016−2020)

GPU code name

Ampere GA102

Ellesmere

Market segment

Desktop

Release date

September 2020

June 2016

Interface

PCIe 4.0 x16

PCIe 3.0 x16

Supplementary power connectors

1x 12-pin

1x 6-pin

Boost clock speed

1710 MHz

1266 MHz

Core clock speed

1450 MHz

1120 MHz

Floating-point performance

5,834 gflops

Manufacturing process technology

8 nm

14 nm

Number of transistors

28,300 million

5,700 million

Pipelines / CUDA cores

8704

2304

Power consumption (TDP)

320 Watt

150 Watt

Texture fill rate

465.1

182.3

Memory type

GDDR6X

GDDR5

Maximum RAM amount

10 GB

8 GB

Memory bus width

320 Bit

256 Bit

Memory clock speed

19000 MHz

8000 MHz

Memory bandwidth

760.3 GB/s

256.0 GB/s

Shared memory

Display Connectors

1x HDMI, 3x DisplayPort

HDMI

DirectX

12 Ultimate (12_2)

12 (12_0)

CUDA

8.5

OpenCL

2.0

OpenGL

4.6

Shader Model

6.5

6.4

Vulkan

1.2

1.2.131