Tesla K20s and Tesla M2075 - video card comparison
Tesla K20s
Tesla M2075
Architecture
Kepler (2012−2018)
Fermi 2.0 (2010−2014)
GPU code name
GK110
GF110
Market segment
Workstation
Workstation
Release date
February 2013
July 2011
Interface
PCIe 2.0 x16
PCIe 2.0 x16
Supplementary power connectors
-
1x 6-pin + 1x 8-pin
Core clock speed
575 MHz
574 MHz
Floating-point performance
3,524 gflops
1,027.7 gflops
Manufacturing process technology
28 nm
40 nm
Number of transistors
7,080 million
3,000 million
Pipelines / CUDA cores
2496
448
Power consumption (TDP)
225 Watt
200 Watt
Texture fill rate
119.6
32.14
Memory type
GDDR5
GDDR5
Maximum RAM amount
5 GB
6 GB
Memory bus width
320 Bit
384 Bit
Memory clock speed
5200 MHz
3132 MHz
Memory bandwidth
208.0 GB/s
150.3 GB/s
Display Connectors
No outputs
No outputs
DirectX
12 (11_0)
12 (11_0)
CUDA
3.5
2.0
OpenCL
1.2
1.1
OpenGL
4.6
4.6
Shader Model
5.1
5.1
Vulkan
1.1.126
N/A