After "a lot of pain" I managed to install all drivers to run CUDA on Ubuntu 16.04. I am having a NVIDIA GeForce 940M graphics adapter. It should have 4GB of memory, but the system informs:
Detected 1 CUDA Capable device(s) Device 0: "GeForce 940M" CUDA Driver Version / Runtime Version 8.0 / 8.0 CUDA Capability Major/Minor version number: 5.0 Total amount of global memory: 2002 MBytes(2099642368 bytes) ( 3) Multiprocessors, (128) CUDA Cores/MP: 384 CUDA Cores GPU Max Clock rate: 1176 MHz (1.18 GHz) Memory Clock rate: 900 Mhz Memory Bus Width: 64-bit L2 Cache Size: 1048576 bytes ...
It says only 2 GB - why? I also have an integrated Intel graphics chip:
$: lspci | grep VGA 00:02.0 VGA compatible controller: Intel Corporation Sky Lake Integrated Graphics (rev 07)
Can I "trick the system" to use only the integrated Intel graphics card, so that I can use the whole power of the NVIDIA GPU in order to run my computations (I am using cuDNN and tensorflow)?
I am not very experienced with these hardware aspects, so I apologize if the question is stupid.