Radeon HD 5670 640SP Edition and GeForce RTX 5080 - video card comparison
Radeon HD 5670 640SP Edition
GeForce RTX 5080
Architecture
TeraScale 2 (2009−2015)
Blackwell 2.0
GPU code name
Juniper
GB203
Market segment
Desktop
Desktop
Release date
July 2010
2025 (recently)
Interface
PCIe 2.0 x16
PCIe 5.0 x16
Supplementary power connectors
None
1x 16-pin
Boost clock speed
-
2520 MHz
Core clock speed
750 MHz
-
Floating-point performance
960.0 gflops
-
Manufacturing process technology
40 nm
0 nm
Number of transistors
1,040 million
-
Pipelines / CUDA cores
640
10752
Power consumption (TDP)
64 Watt
350 Watt
Texture fill rate
24.00
846.7
Memory type
GDDR5
GDDR7
Maximum RAM amount
512 MB
16 GB
Memory bus width
128 Bit
256 Bit
Memory clock speed
4000 MHz
5 GB/s
Memory bandwidth
64 GB/s
160.0 GB/s
Display Connectors
1x DVI, 1x HDMI, 1x DisplayPort
1x HDMI 2.1, 3x DisplayPort 1.4a
HDMI
+
+
DirectX
11.2 (11_0)
12 Ultimate (12_2)
CUDA
-
9.1
OpenCL
1.2
3.0
OpenGL
4.4
4.6
Shader Model
5.0
6.7
Vulkan
N/A
1.3