MATLAB Answers

0

Why is arrayfun for GPU slower than normal operations

Asked by Theron FARRELL on 28 May 2019
Latest activity Commented on by Theron FARRELL on 3 Jun 2019
Hi there,
Here goes a piece of testing code, yet arrayfun runs more slowly. Any thoughts? Many thanks.
function Test_GPU1()
EP = gpuArray(eps*ones(10000, 1, 'single'));
ONE = gpuArray(ones(10000, 1, 'single'));
ZERO = gpuArray(zeros(10000, 1, 'single'));
Cur_FF_Output = gpuArray(0.5*ones(10000, 1, 'single'));
Cur_Desired_Output = gpuArray(0.5*ones(10000, 1, 'single'));
for iter = 1:1000
% In output layer, Cur_Delta = Del(C)/Del(z) = Del(C)/Del(a) * Del(a)/Del(z)
% [~, Cur_Delta0] = Cost_Function_GPU(Cur_FF_Output, Cur_Desired_Output, Hyper_Para);
temp00 = Cur_FF_Output + eps;
temp11 = log(temp00);
temp22 = log(1-Cur_FF_Output+eps);
temp33 = Cur_Desired_Output.*temp11;
temp44 = 1-Cur_FF_Output.*temp22;
Cur_Delta = Cur_FF_Output-Cur_Desired_Output;
Cost = 0-sum(temp33+temp44);
temp00 = arrayfun(@plus, Cur_FF_Output, EP);
temp11 = arrayfun(@log, temp00);
temp22 = arrayfun(@log, arrayfun(@minus, ONE, arrayfun(@plus, Cur_FF_Output, EP)));
temp33 = arrayfun(@times, Cur_Desired_Output, temp11);
temp44 = arrayfun(@minus, ONE, arrayfun(@times, Cur_FF_Output, temp22));
Cur_Delta = arrayfun(@minus, Cur_FF_Output, Cur_Desired_Output);
Cost = arrayfun(@minus, ZERO, sum(temp33+temp44));
end
end

  1 Comment

Of course arrayfun has a certain overhead. It is expected to run slower than calling the operators directly.

Sign in to comment.

2 Answers

Answer by Joss Knight
on 28 May 2019
 Accepted Answer

You are misunderstanding the use of arrayfun for gpuArray. Combine all those operations into a single function.
temp00 = arrayfun(@plus, Cur_FF_Output, EP);
temp11 = arrayfun(@log, temp00);
temp22 = arrayfun(@log, arrayfun(@minus, ONE, arrayfun(@plus, Cur_FF_Output, EP)));
temp33 = arrayfun(@times, Cur_Desired_Output, temp11);
temp44 = arrayfun(@minus, ONE, arrayfun(@times, Cur_FF_Output, temp22));
Cur_Delta = arrayfun(@minus, Cur_FF_Output, Cur_Desired_Output);
Cost = arrayfun(@minus, ZERO, sum(temp33+temp44));
becomes
function Cur_Delta = stuff(Cur_FF_Output, Cur_Desired_Output, EP)
temp00 = Cur_FF_Output + EP;
temp11 = log(temp00);
temp22 = log(1 - (Cur_FF_Output + EP));
temp33 = Cur_Desired_Output .* temp11;
temp44 = 1 - (Cur_FF_Output .* temp22);
Cur_Delta = Cur_FF_Output - Cur_Desired_Output;
end
Cur_Delta = arrayfun(@stuff, Cur_FF_Output, Cur_Desired_Output, EP);
Obviously, this can be extremely simplified. I've made a start, by removing the unnecessary ONE and ZERO variables.
After this, question whether you really need arrayfun, or should just call this function directly? MATLAB uses some clever optimisations that, for most sequences of element-wise operations, make using arrayfun unnecessary.

  7 Comments

I see. 'If you know what you're doing', 'At fast as', and 'no fun' are the key phrases, as I sense. *_^
To be candid, MATLAB, being the de facto most powerful, miraculous, as well as user-friendly scientific and technical SIMULATION and PROTOTYPING tool since 1984--I would not use the word computing here (let's forget MATLAB coder, embedded coder etc originally designed for auto industry at the moment), enjoys her AUTOMATIC optimisation without users' heavy involvement. One of the most typical examples is element-wise (vectorised) operations. That being said, my point is that a user should not take too many efforts on seeking THE most optimised code in lieu of concentrating on algorithmic designs and prototyping, which I do not think most users will do. Consequently, some notifications about pros and cons of using functions as well as PRACTICAL examples such as arrayfun() would be better to be given in the Help page, for example a more formal statement of your words above. Especially , in the advent of DNN, MATLAB would be better prepared for competing with loads of well-optimised, open-sourced code, tensorflow, theano, Caffe etc...
I don't know how recently you viewed the latest documentation on GPU support. I think it's pretty comprehensive, and doesn't encourage you to use arrayfun unnecessarily. There's a great page that talks you through the various options for optimising your code which mentions arrayfun only as an advanced manoeuvre that might, but won't always, improve your performance. Generally we think of arrayfun as the way to 'write custom kernels in the MATLAB language', which many advanced users may look for. We rarely find it useful to document exactly what optimizations MATLAB is using. Usually it just confuses people, who start looking for performance improvements in the wrong place, or start blaming unexpected behaviour on the optimisations. Best to just do our best and leave ourselves the wiggle room to change the way things work from one release to the next.
Understood! Thanks again for your great help and detailed explanation. I am always patient with MATLAB since 1997:-)

Sign in to comment.


Answer by Jan
on 28 May 2019
Edited by Jan
on 28 May 2019

Of course arrayfun has a certain overhead. It is expected to run slower than calling the operators directly with arrays as inputs. In addition, in
Cur_FF_Output + eps
the second operand is a scalar, while in
arrayfun(@plus, Cur_FF_Output, EP)
Matlab has to process a vector. Addressing the elements of an array needs to access memory using a loop. Accessing a scalar is much cheaper.
What is the purpose of:
arrayfun(@minus, ZERO, sum(temp33+temp44))
? This is faster:
-sum(temp33+temp44)

  4 Comments

Show 1 older comment
If that is true, it is a bug. Are you sure that sum(temp33+temp44) is a gpuArray? What happens if you use gpuArray(0) instead?
No, it is not a bug. Maybe somewhere I wrote the code mistakenly. My bad, sorry.
Even arrayfun(@minus, 0, sum(temp33+temp44)) is too complicated compared to
-sum(temp33+temp44)

Sign in to comment.