On Mon, Oct 30, 2006 at 02:45:38AM +0000, david eddy wrote: > I believe this is the reason the cutoff for the exponent ranges is so clear > cut: > a small change in exponent size produces a dramatic change in the probabiliy > of a rounding error>0.5
I have no idea why you think these ranges are the only true ones and drop automatically from nowhere. Prime95 ran with completely different ranges, picked more or less out of the thin air, for ages, before an error analysis was done and the current ranges set. Heck, if you want to test an exponent near the limit, Prime95 first runs a couple hundred iterations, measures the rounding error and picks the FFT size from that. /* Steinar */ -- Homepage: http://www.sesse.net/ _______________________________________________ Prime mailing list [email protected] http://hogranch.com/mailman/listinfo/prime
