> Assume that a set of 400,000 exponents get single checked,
> and double checked, and the error rate per check is 0.5%.
> If the error occurrence is independent, that means about 4000
> will not match. Of these 4000 then the triple checks would
> have errors in about 20, and require a quadruple check which
> most likely would match one of the previous results.
But, just to clarify, even if both a first time check and a double check
both fail and come up with a bad residue for some reason, the partial
residues from each run won't match, so they'd still be caught and flagged
for a triple check. A triple check will match the first or the second run,
or maybe neither requiring a 4th run.
It all sounds pretty solid to me. :-)
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
