Why is pagemtimes slower than just coding up the matrix multiplica​tion?Espec​ially on GPU.

62 visualizaciones (últimos 30 días)
I'm going to use the Pagemtimes function in my custom loss function. But when I train my network with GPU, it doesn't work very well. I found some people asking questions about this in the community, but there wasn't an answer that could be taken on board. Here are my tests for an examples of questions already in the community.
function C=pagemtimes_version(A,B,E,F)
C = pagemtimes(F,(B+pagemtimes(E,A)));
end
function C=direct(A,B,E,F)
C(:,:,1,1) = ...
F(:,:,1,1).*(A(:,:,1,1).*E(:,:,1,1)+B(:,:,1,1)) +...
F(:,:,1,2).*(A(:,:,1,1).*E(:,:,1,2)+B(:,:,1,1)) +...
F(:,:,1,3).*(A(:,:,1,1).*E(:,:,1,3)+B(:,:,1,1));
C(:,:,2,1) = ...
F(:,:,2,1).*(A(:,:,2,1).*E(:,:,2,1)+B(:,:,2,1)) +...
F(:,:,2,2).*(A(:,:,2,1).*E(:,:,2,2)+B(:,:,2,1)) +...
F(:,:,2,3).*(A(:,:,2,1).*E(:,:,2,3)+B(:,:,2,1));
C(:,:,3,1) = ...
F(:,:,3,1).*(A(:,:,3,1).*E(:,:,3,1)+B(:,:,3,1)) +...
F(:,:,3,2).*(A(:,:,3,1).*E(:,:,3,2)+B(:,:,3,1)) +...
F(:,:,3,3).*(A(:,:,3,1).*E(:,:,3,3)+B(:,:,3,1));
end
Since some of the replies suggested a single-precision test, I'll show it in single-precision first.
Nx=1000;
Ny=1000;
[E,F] = deal(gpuArray(single(rand(Nx,Ny,3,3))));
[A,B] = deal(gpuArray(single(rand(Nx,Ny,3,1))));
timeit(@()direct(A,B,E,F))
ans = 2.9201e-04
timeit(@()pagemtimes_version(A,B,E,F))
ans = 0.0045
The difference is almost 20 times, and the larger the array the greater the difference in effect, when Nx,Ny takes 5000 the difference is 1000 times (0.1/10^-4)
[E,F] = deal(single(rand(Nx,Ny,3,3)));
[A,B] = deal(single(rand(Nx,Ny,3,1)));
timeit(@()direct(A,B,E,F))
ans = 0.0421
timeit(@()pagemtimes_version(A,B,E,F))
ans = 0.0514
GPU even slower than CPU in double-precision.
[E,F] = deal(gpuArray(rand(Nx,Ny,3,3)));
[A,B] = deal(gpuArray(rand(Nx,Ny,3,1)));
timeit(@()direct(A,B,E,F))
ans = 2.6526e-04
timeit(@()pagemtimes_version(A,B,E,F))
ans = 0.1517
[E,F] = deal(rand(Nx,Ny,3,3));
[A,B] = deal(rand(Nx,Ny,3,1));
timeit(@()direct(A,B,E,F))
ans = 0.0874
timeit(@()pagemtimes_version(A,B,E,F))
ans = 0.1163
Pagemtimes are really handy. But it doesn't look good for double precision data and on the GPU. I would like to know if there is any way to fix
  4 comentarios
埃博拉酱
埃博拉酱 el 1 de Nov. de 2024 a las 1:20
Please accept the best answer if your question is solved.

Iniciar sesión para comentar.

Respuestas (2)

Joss Knight
Joss Knight el 31 de Oct. de 2024 a las 15:54
Your implementation is incorrect I'm afraid, you are using elementwise times rather than mtimes. You are also using timeit instead of gputimeit which is unfairly penalizing the pagemtimes code because it is running synchronously.
  1 comentario
Hongbo Sun
Hongbo Sun el 31 de Oct. de 2024 a las 21:34
Editada: Hongbo Sun el 31 de Oct. de 2024 a las 21:57
Thank you. I corrected the code in the example, and after using gputimeit correctly, everything worked.I'll look elsewhere to speed up my program.

Iniciar sesión para comentar.


the cyclist
the cyclist el 31 de Oct. de 2024 a las 15:01
It seems to me that the two functions are not calculating the same thing, based on the size of their respective outputs:
rng default
Nx=1000;
Ny=1000;
[E,F] = deal(single(rand(Nx,Ny,3,3)));
[A,B] = deal(single(rand(Nx,Ny,3,1)));
C1 = pagemtimes_version(A,B,E,F);
C2 = direct(A,B,E,F);
size(C1)
ans = 1×4
1000 1000 3 3
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
size(C2)
ans = 1×3
1000 1000 3
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
function C=pagemtimes_version(A,B,E,F)
C = pagemtimes(F,(B+pagemtimes(E,A)));
end
function C=direct(A,B,E,F)
C(:,:,1,1) = ...
F(:,:,1,1).*(A(:,:,1,1).*E(:,:,1,1)+B(:,:,1,1)) +...
F(:,:,1,2).*(A(:,:,1,1).*E(:,:,1,2)+B(:,:,1,1)) +...
F(:,:,1,3).*(A(:,:,1,1).*E(:,:,1,3)+B(:,:,1,1));
C(:,:,2,1) = ...
F(:,:,2,1).*(A(:,:,2,1).*E(:,:,2,1)+B(:,:,2,1)) +...
F(:,:,2,2).*(A(:,:,2,1).*E(:,:,2,2)+B(:,:,2,1)) +...
F(:,:,2,3).*(A(:,:,2,1).*E(:,:,2,3)+B(:,:,2,1));
C(:,:,3,1) = ...
F(:,:,3,1).*(A(:,:,3,1).*E(:,:,3,1)+B(:,:,3,1)) +...
F(:,:,3,2).*(A(:,:,3,1).*E(:,:,3,2)+B(:,:,3,1)) +...
F(:,:,3,3).*(A(:,:,3,1).*E(:,:,3,3)+B(:,:,3,1));
end
  1 comentario
Hongbo Sun
Hongbo Sun el 31 de Oct. de 2024 a las 21:13
Editada: Hongbo Sun el 31 de Oct. de 2024 a las 22:02
You're right, I used the example from the previous question but didn't think it through. But the main problem is the incorrect use of the timeit function.

Iniciar sesión para comentar.

Productos


Versión

R2024a

Community Treasure Hunt

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

Start Hunting!

Translated by