GRID M40 and Tesla P40 - video card comparison
GRID M40
Tesla P40
Architecture
Maxwell (2014−2018)
Pascal (2016−2021)
GPU code name
GM107
GP102
Market segment
Workstation
Workstation
Release date
May 2016
September 2016
Interface
PCIe 3.0 x16
PCIe 3.0 x16
Supplementary power connectors
-
8-pin EPS
Boost clock speed
1000 MHz
1531 MHz
Core clock speed
1033 MHz
1303 MHz
Floating-point performance
-
11,758 gflops
Manufacturing process technology
28 nm
16 nm
Number of transistors
1,870 million
11,800 million
Pipelines / CUDA cores
384
3840
Power consumption (TDP)
50 Watt
250 Watt
Texture fill rate
33.06
367.4
Memory type
GDDR5
GDDR5
Maximum RAM amount
8 GB
24 GB
Memory bus width
128 Bit
384 Bit
Memory clock speed
5200 MHz
7.2 GB/s
Memory bandwidth
83.2 GB/s
347.1 GB/s
Display Connectors
No outputs
No outputs
DirectX
12 (11_0)
12 (12_1)
CUDA
5.0
6.1
OpenCL
1.2
3.0
OpenGL
4.6
4.6
Shader Model
5.1
6.7
Vulkan
1.1.126
1.3