That's what threads are good for, a performance boost. But I believe I can get squash to run on the strong arm just fine.

(Right now, the most significant overhead from moving away from threads will be the spectrum analyser, which isn't even present in the current empeg version of squash. And I think that I could used shared memory or think of something, if performance becomes a problem -- we'll build that bridge when we get there).

I'm mainly concerned with the influence threads has on the design. I think I can get a good multi-programming design going that may actually load the processor less (because I think I will have traded off locking overhead for less IPC overhead).

Anyway, wish me luck; right now things aren't looking good as my car broke down and so my radio is with it in another city.