As requested by David, a verification of the new Milton Brown-Mills
heuristic for a prime search of the form 12345*2^n +1 . (Not a GFN
form, as stated). Courtesy of NewPGen (Paul Jobling) and Proth (Yves
Gallot) and comparative statistical analysis by David Broadhurst
(forthcoming). Primers are invited to search for the `Pineapple
Rag' midi file by Scott Joplin on the net in celebration and to
generally `have a good day!'
12345*2^29100 + 1 is prime ! (a = 7) [8765 digits]
12345*2^29100 + 1 is prime ! (verification : a = 19) [8765 digits]
- Paul Mills announced:
> The Brown-Mills heuristic algorithm is 10Xand adduced the example:
> faster than any current prime searching algorithm.
> 12345*2^29100 + 1 is prime ! (a = 7)I looked to see how much CPUtime such small primes take
were found in 6 hours at 1GHz.
So Paul, all you have to do
is generate a hundred 30kbit primes
each CPU day, and we shall be convinced.
- On Sun, 03 February 2002, "djbroadhurst" wrote:
> Paul Mills announced:Adduced, eh? Nice word, but I think I'll stick with 'presented' 'cited', or 'proffered'.
> > The Brown-Mills heuristic algorithm is 10X
> > faster than any current prime searching algorithm.
> and adduced the example:
> > 12345*2^29100 + 1 is prime ! (a = 7)
However, the gap upon which noone has yet pounced is the 'current prime searching algorithms'. What algorithms? I've seen more convincing soap powder "better than all others" spiel(s).
> I looked to see how much CPUtime such small primes takeMaybe you didn't use a 'current' prime searching algorithm??
> were found in 6 hours at 1GHz.
> So Paul, all you have to do
> is generate a hundred 30kbit primes
> each CPU day, and we shall be convinced.
Don't be fooled, CRC Press are _not_ the good guys.
They've taken Wolfram's money - _don't_ give them yours.
Find the best deals on the web at AltaVista Shopping!