Radeon HD 3850 and GeForce 8800 GTX - video card comparison
Passmark - benchmark score Radeon HD 3850 vs GeForce 8800 GTX
Radeon HD 3850
404
GeForce 8800 GTX
608
Radeon HD 3850
GeForce 8800 GTX
Architecture
TeraScale (2005−2013)
Tesla (2006−2010)
GPU code name
RV670
G80
Market segment
Desktop
Desktop
Release date
November 2007
November 2006
Interface
PCIe 2.0 x16
PCIe 1.0 x16
Supplementary power connectors
1x 6-pin
2x 6-pin
SLI options
-
+
Core clock speed
668 MHz
576 MHz
CUDA cores
-
575
Floating-point performance
427.5 gflops
345.6 gflops
Manufacturing process technology
55 nm
90 nm
Number of transistors
666 million
681 million
Pipelines / CUDA cores
320
128
Power consumption (TDP)
75 Watt
155 Watt
Texture fill rate
10.69
36.8 billion/sec
Memory type
GDDR3
GDDR3
Maximum RAM amount
1 GB
768 MB
Memory bus width
256 Bit
384 Bit
Memory clock speed
1660 MHz
900 MHz
Memory bandwidth
53.12 GB/s
86.4 GB/s
Display Connectors
2x DVI, 1x S-Video
2x DVI, 1x S-Video
DirectX
10.1 (10_1)
11.1 (10_0)
CUDA
-
+
OpenCL
N/A
1.1
OpenGL
3.3
3.3
Shader Model
4.1
4.0
Vulkan
N/A
N/A