I just ran the same tests (though with smaller arrays of 25k and the mach timing) on an iPhone 4S running iOS 5 and found that vvsin and vvsqrt were both consistently better than a regular sin or sqrt loop for the entire range of array sizes (roughly 30-40% better speeds).
Interesting that we see the problems on the Xeon side, but the A5 chips seem to have consistently better results across the board.
Do not post admin requests to the list. They will be ignored.
PerfOptimization-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden