Code Optimization

[ent1ty]

Hi, I've found quite a nice site about optimization of code. I have not read it all yet, but so far it looks good.
http://www.azillionmonkeys.com/qed/optimize.html

[Bate]

thanks, looks promising. I'm gonna read it soon

The article is from 2007 though, not sure if modern compilers like VS2008 or 2010 are able to optimize all of that code automatically.

[slavik262]

It's an interesting article, but I don't know how valid it is. Some of it seems like unnecessary obfuscation that might not even produce the intended results. Other things like switching x * 8 to x << 3, are done automatically by any compiler worth its salt (every modern compiler I can think of knows to automatically make multiplication with powers of two bit shift ops automatically). "Optimizing" code for certain hardware also seems silly to me too - hardware changes, and it changes fast. What's true today probably won't be true a year from now, and it seems to me like you'd be wasting a lot of time doing these "optimizations" instead of focusing on what your code is supposed to be doing. Well written and thought out code will always outperform poorly written code that's "optimized" the way this article describes. In short, let the optimizing compiler do the optimization and focus on making your code good.

[randomMesh]

Thinking in C++, 2nd ed. Volume 1, Programming Guidelines

1. First make it work, then make it fast. This is true even if you are certain that a piece of code is really important and that it will be a principal bottleneck in your system. Don’t do it. Get the system going first with as simple a design as possible. Then if it isn’t going fast enough, profile it. You’ll almost always discover that “your” bottleneck isn’t the problem. Save your time for the really important stuff.

35. Don’t fall prey to premature optimization. That way lies madness. In particular, don’t worry about writing (or avoiding) inline functions, making some functions nonvirtual, or tweaking code to be efficient when you are first constructing the system. Your primary goal should be to prove the design, unless the design requires a certain efficiency.

[agamemnus]

1. First make it work, then make it fast.

Agreed, but not.

Balance is important-- not just in programming but in life, too...

Making things "work" without making them reasonably fast will lead to a horrible mess; making things insanely fast before making them work will lead to nothing.

[slavik262]

I know several computer engineers and a professional game developer, and every time I talk to them about optimization, they say that it's best to let the compiler do the simple optimizations (such as knowing whether to bit shift or multiply or divide) and focus on optimizing your program at a higher level. For example, no matter what compiler you use, heapsort will (almost always) outperform bubblesort. Work on the higher-level structure of your code and let the optimizing compiler deal with the details.