Radeon HD 4860 and GeForce RTX 3070 Ti - video card comparison
Radeon HD 4860
GeForce RTX 3070 Ti
Architecture
TeraScale (2005−2013)
Ampere (2020−2022)
GPU code name
RV790
Ampere GA104
Market segment
Desktop
Desktop
Release date
September 2009
May 2021
Interface
PCIe 2.0 x16
PCIe 4.0 x16
Supplementary power connectors
2x 6-pin
1x 12-pin
Boost clock speed
-
1770 MHz
Core clock speed
700 MHz
1575 MHz
Floating-point performance
896.0 gflops
-
Manufacturing process technology
55 nm
8 nm
Number of transistors
959 million
17,400 million
Pipelines / CUDA cores
640
6144
Power consumption (TDP)
130 Watt
290 Watt
Texture fill rate
22.40
339.8
Memory type
GDDR5
GDDR6X
Maximum RAM amount
512 MB
8 GB
Memory bus width
256 Bit
256 Bit
Memory clock speed
3000 MHz
19000 MHz
Memory bandwidth
96 GB/s
608.3 GB/s
Shared memory
-
-
Display Connectors
2x DVI, 1x S-Video
1x HDMI, 3x DisplayPort
HDMI
-
+
DirectX
10.1 (10_1)
12 Ultimate (12_2)
CUDA
-
8.6
OpenCL
1.1
3.0
OpenGL
3.3
4.6
Shader Model
4.1
6.6
Vulkan
N/A
1.2