Paul Jones has taken a look at some performance statistics that Rasmus Lerdorf shared in a recent presentation from some of his tests.

It was with some interest, then, that I viewed Rasmus Lerdorf's slides on the subject of performance benchmarking. I'm beginning to think there's something unexpected or unexamined in his testing methodology.

Paul compares some of his statistics (gathered from his testing) to those of Rasmus and finds a few places where things don't quite match up - specifically in the default overhead that seems to be a part of Rasmus' stats. Paul goes on to look at EC2 as a cause to the problems but ends up without much of a confirmed answer.