Radeon RX 580 and Data Center GPU Max Subsystem - video card comparison

Radeon RX 580

Data Center GPU Max Subsystem

Architecture

Polaris (2016−2019)

Generation 12.5 (2021−2022)

GPU code name

Polaris 20

Ponte Vecchio

Market segment

Desktop

Release date

April 2017

January 2023

Interface

PCIe 3.0 x16

PCIe 5.0 x16

Supplementary power connectors

1x 8-pin

1x 16-pin

Boost clock speed

1340 MHz

1600 MHz

Core clock speed

1257 MHz

Floating-point performance

6,175 gflops

Manufacturing process technology

14 nm

10 nm

Number of transistors

5,700 million

100,000 million

Pipelines / CUDA cores

2304

16384

Power consumption (TDP)

185 Watt

2400 Watt

Texture fill rate

193.0

1,638

Memory type

GDDR5

HBM2e

Maximum RAM amount

8 GB

128 GB

Memory bus width

256 Bit

8192 Bit

Memory clock speed

8000 MHz

3.1 GB/s

Memory bandwidth

256.0 GB/s

3,205 GB/s

Display Connectors

1x HDMI, 3x DisplayPort

No outputs

HDMI

DirectX

12 (12_0)

12 (12_1)

OpenCL

2.0

3.0

OpenGL

4.6

Shader Model

6.4

6.6

Vulkan

1.2.131

N/A