Hi all,
As promised, the server now has about 1000 small exponents
to give out for triple-checks.
Also, the server has about 1000 new exponents between 7 million and
12 million to assign for first-time tests. These exponents were tested once
but had at least one ROUNDOFF > 0.4 or SUM(INPUTS) != SUM(OUTPUTS)
error. These LL tests have at best a 50% chance of being correct. Rather
than wait for double-checking to reach these lofty levels, I think it best to
reassign them for first-time tests.
I just finished coding and debugging the 2K FFT (exponents up to 44000)
for the P4. The old code took 323000 clocks, the new code takes 117000.
That's 2.75 times faster -- I'll see if I can't do a little better....
-- George
_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
