Weird context-switching performance

[Milo Hyson]

We've got these two very similar machines that are exhibiting
drastically different context-switching performance. The slower of the
two is actually task-switching at four times the rate of the other.
We've gone through several benchmarks and are at a complete loss to
explain it. We've set up a page documenting our efforts:

http://www.cyberlifelabs.com/cs-benchmark/

I would appreciate it if someone could offer some insight into what
affects context-switching. Is there some obscure kernel variable that
we've overlooked? Or is it just one of those unexplainable feng-shui
situations?

Thanks in advance.

-- 
Milo Hyson
Chief "Mad" Scientist
CyberLife Labs, LLC