Radeon HD 3850 and GeForce GTX 1650 GDDR6 - video card comparison
Radeon HD 3850
GeForce GTX 1650 GDDR6
Architecture
TeraScale (2005−2013)
Turing (2018−2021)
GPU code name
RV670
TU117
Market segment
Desktop
Desktop
Release date
November 2007
April 2020
Interface
PCIe 2.0 x16
PCIe 3.0 x16
Supplementary power connectors
1x 6-pin
None
Boost clock speed
-
1590 MHz
Core clock speed
668 MHz
-
Floating-point performance
427.5 gflops
-
Manufacturing process technology
55 nm
12 nm
Number of transistors
666 million
4,700 million
Pipelines / CUDA cores
320
896
Power consumption (TDP)
75 Watt
75 Watt
Texture fill rate
10.69
89.04
Memory type
GDDR3
GDDR6
Maximum RAM amount
1 GB
4 GB
Memory bus width
256 Bit
128 Bit
Memory clock speed
1660 MHz
12000 MHz
Memory bandwidth
53.12 GB/s
192.0 GB/s
Display Connectors
2x DVI, 1x S-Video
1x DVI, 1x HDMI, 1x DisplayPort
HDMI
-
+
DirectX
10.1 (10_1)
12 (12_1)
CUDA
-
7.5
OpenCL
N/A
1.2
OpenGL
3.3
4.6
Shader Model
4.1
6.5
Vulkan
N/A
1.2.140