Lucas Wiman on Thu, 14 Oct 1999 17:17:22 -0400 (EDT)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

SPRP's, and miller(n,k) 

Hi,

In the source code for PARI, it lists the result:
 * Moreover, the four bases chosen at
 *
 * k == 16  (2,13,23,1662803)  will correctly detect all composites up
 * to at least 10^12, and the combination at
for the function miller().  Using Richard Pinch's list of strong
psuedoprimes base 2, up to 10^13, I found that this is true for
n<1122004669633=611557*1834669.

Also, for 2,11,13,23,1662803 correctly detects all composites up
to 10^13 (and possibly higher).  This is better than the successive
primes result (for 5 Miller SPRP tests) by a factor of 4.6.

Lucas Wiman