Sunday, January 17, 2010

Objects vs. the L2 Cache

An excellent little performance optimization presentation that shows how important memory layout is for today's processors:

The beginning of the talk makes the observation that since C++ was started in 1979 the cost of accessing uncached main memory has ballooned from 1 cycle to 400 cycles.

The bulk of the presentation shows the optimization of a graphics hierarchy library, where switching from a traditional OO design to a structure-of-arrays design makes the code run 3 times faster. (Because of better cache access patterns.)

1 comment:

Alan Mimms said...

Hi Jack, remember me from Apple?

I found this very interesting. It didn't have a lot that was NEW, per se, but it was all in one place and well presented. I am doing high performance networking stuff at work, and high performance Android game development at home.

It's nice to see you're still as energetic as ever!