At 08:55 AM 10/24/2006, david eddy wrote: >I'm mystified. > >The number of factors of 2^p - 1 to try up to 2^m goes as (2^m)/p >Why don't the times follow suit?
Both are correct. You are right there are twice as many factors to test. However, the benchmark times how long it takes to test a small fixed size subset of factors of that size. >The only feature I understand is the jump at 64 bits, since the division >suddenly gets harder. Again you are correct. _______________________________________________ Prime mailing list [email protected] http://hogranch.com/mailman/listinfo/prime
