> Really we need only consider IEEE single (24+8) & double (53+11) precision > types... the x87 80-bit format is not much different to double precision the > way we use it, and I'm not aware of any common hardware implementations of > other floating-point formats.
At the risk of sounding dumb (Always a problem for me!) can we not just add up the iterations as a total? Sure different FFT ranges are more efficiently completed than other FFT ranges, but the end result is exponents checked, right? So a 5M exponent would be worth 1/2 of a 10M exponent... This would be like saying "I went to South Carolina last week"... Who cares if I walked or took a plane? The end result is I got there... And if Florida is 2x farther, then that is how it should be counted, not in Yugo-hours versus 747-hours... Or am I missing something? _________________________________________________________________________ Unsubscribe & list info -- http://www.ndatech.com/mersenne/signup.htm Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
