Breaking up a computation vs "..." to continue line - Huge Speed Difference !?

Question

Dan Ryan 2013-2-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/63608-breaking-up-a-computation-vs-to-continue-line-huge-speed-difference

I was profiling an ode solver today and found something very strange. The system has 9 states and the computation of the ode RHS was taking about 100 times longer to compute the 5th state than any other state.

The 5th state equation was longer than the other equations, but not so drastically longer that it should take 100 times longer to compute. It ran over 4 lines in the editor so there were "..." continuations at the end of each line.

I rewrote the equation using multiple assignments instead of one big assignment and got about a 50 times speed up. By this, I mean I changed

value(5, :) = big_expression1 ...
              + big_expression2 ...
              + big_expression3 ...
              + big_expression4;

to

value(5,:) = big_expression1;
value(5,:) = value(5,:) + big_expression2;
value(5,:) = value(5,:) + big_expression3;
value(5,:) = value(5,:) + big_expression4;

Can you think of any reason why the second version should drastically outperform the first?

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Jan 2013-2-15

It depends on the real processing in "big_expression". Therefore your paraphrased code does not allow to reveal the problem.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

José-Luis 2013-2-15

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/63608-breaking-up-a-computation-vs-to-continue-line-huge-speed-difference#answer_75209

编辑：José-Luis 2013-2-15

I don't think it has anything to do with "...". Shot in the dark:

Cache locality? Maybe big_expression(n) and big_expression(n+1) fit together in the cache whereas when you call them all at the same time values need to be moved around in memory.