GeForce GTX 680 and GeForce GTX 580 - video card comparison
Passmark - benchmark score GeForce GTX 680 vs GeForce GTX 580
GeForce GTX 680
5721
GeForce GTX 580
4702
GeForce GTX 680
GeForce GTX 580
Architecture
Kepler (2012−2018)
Fermi (2010−2014)
GPU code name
GK104
GF110
Market segment
Desktop
Desktop
Release date
March 2012
November 2010
Bus support
PCI Express 3.0
PCI-E 2.0 x 16
Interface
PCIe 3.0 x16
PCIe 2.0 x16
Supplementary power connectors
Two 6-pin
One 6-pin and One 8-pin
SLI options
+
+
Boost clock speed
1058 MHz
-
Core clock speed
1006 MHz
772 MHz
CUDA cores
1536
512
Floating-point performance
3,090.4 gflops
1,581.1 gflops
Manufacturing process technology
28 nm
40 nm
Number of transistors
3,540 million
3,000 million
Pipelines / CUDA cores
1536
512
Power consumption (TDP)
195 Watt
244 Watt
Maximum GPU temperature
-
97 °C
Texture fill rate
128.8 billion/sec
49.4 billion/sec
Memory type
GDDR5
GDDR5
Maximum RAM amount
2048 MB
1536 MB
Memory bus width
256-bit GDDR5
384 Bit
Memory clock speed
6000 MHz
2004 MHz (4008 data rate)
Memory bandwidth
192.2 GB/s
192.4 GB/s
Shared memory
-
-
Display Connectors
One Dual Link DVI-I, One Dual Link DVI-D, One HDMI, One DisplayPort
Mini HDMITwo Dual Link DVI
Multi monitor support
4 displays
+
HDMI
+
+
HDCP
+
-
Maximum VGA resolution
2048x1536
2048x1536
Audio input for HDMI
Internal
Internal
DirectX
12 (11_0)
12 (11_0)
CUDA
+
+
OpenCL
1.2
1.1
OpenGL
4.2
4.2
Shader Model
5.1
5.1
Vulkan
1.1.126
+