I am trying to enable the GPU running in my cmdstanr. My laptop has both Intel and NVIDIA GPU for windows. I tried to follow the instructions that is given by @Erik_Strumbelj
This is my frist time that ia trying to do these steps and I am not famliar with OpenCL. I tried to google and find solutions but could not find good approach:
1. how can i check or install the Make mingw32-make and g++ 4.9.3? I have installed the cmdstanr from anaconda. when I use the conda commander and checks them it gives me:
(stan-env) C:\Users\mmabae>g++ --version g++ (Rev5, Built by MSYS2 project) 5.3.0 Copyright (C) 2015 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
(stan-env) C:\Users\mmabae>mingw32-make --version GNU Make 4.1.90 Built for Windows32 Copyright (C) 1988-2014 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law.
so does it mean everything is ok for this step? having said that I already installed NVIDIA CUDA and Rtools as well.
I have no idea how i should check the IDs of my target device. I already instaled visual studio and when i use âmakeâ there it said this word is not meaning. i do not about clifinfo as well.
I will be thankfull if you kindly advice me on this.
The post says âWhile modern Intel GPUs support OpenCL, their double-precision performance is unlikely to bring speedup compared to a modern CPUâ. So this means that I shouldnât expect much acceleration from GPU?
I ask this before embarking in installing the cmdstan for GPU. (I have already my GPU in the desktop, but also 32 CPUs threads, hence I wonder if is it worst to go on with the GPUs).