清空筛选条件
清空筛选条件

How do I make this code GPU capable

3 次查看（过去 30 天）
Mark Storrings
Mark Storrings 2017-11-15
回答： Tony Mohan Varghese 2017-11-29
We are trying to take advantage of the GPU for increase processing speed. As a test, we are trying to convert the following code to utilize the GPU:
clear all; close all; clc;
variables =1000; data = 60;
X = rand(variables, data);
Y = rand(data, 1);
distcomp.feature( 'LocalUseMpiexec', false ); delete(gcp('nocreate')); parpool(4); pctRunOnAll warning('off','all')
disp('CPU started'); tic;
parfor j= [1:variables]
mdl = fitlm( X([ j ],:) , Y , 'linear','RobustOpts','off');
Rsq(j) = mdl.Rsquared.Adjusted;
Rmse(j) =mdl.RMSE;
end
toc
disp('CPU ended');
Any assistance would be greatly appreciated.
  1 个评论
Matt J
Matt J 2017-11-16
编辑：Matt J 2017-11-16
It doesn't look to me like it would benefit from the GPU. Parfor is a good choice for the operations that you've shown.

请先登录，再进行评论。

请先登录，再回答此问题。

回答（1 个）

Tony Mohan Varghese
Tony Mohan Varghese 2017-11-29
Refer to GPU Capabilities and performance in MATLAB Documentation :
https://www.mathworks.com/help/distcomp/gpu-capabilities-and-performance.html
There are built in functions that you can use to improve your code :
https://www.mathworks.com/help/distcomp/run-built-in-functions-on-a-gpu.html

类别

在 Help Center 和 File Exchange 中查找有关 GPU Computing 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by