![]() |
|
|
#1 |
|
Mar 2004
22·33·5 Posts |
I was just wondering, is the cost/benefit calculation for how much factoring gets done to a Mersenne number based on the assumption that the number will be LL tested twice, or just once?
|
|
|
|
|
|
#2 |
|
Oct 2007
Manchester, UK
22×3×113 Posts |
Twice.
Normally I'd link you to the maths page on mersenne.org now, but looking up, copying and pasting links on a phone takes more than the tenth of a second it would take on a real PC. I'm sure you can find it short order anyway, it's in one of the lower down sidebar sections I believe. |
|
|
|
|
|
#3 | |
|
Jun 2003
2·3·7·112 Posts |
Quote:
EDIT:- Probably I'm wrong about the 2.032, but I seem to recall such a figure having some significance. Last fiddled with by axn on 2010-11-16 at 16:32 |
|
|
|
|
|
|
#4 |
|
Nov 2010
Ann Arbor, MI
2·47 Posts |
This is the formula used (Last Updated: September 1, 2009):
factoring_cost < chance_of_finding_factor * 2 * primality_test_cost You can find the whole explanation in the following link: http://www.mersenne.org/various/math.php Last fiddled with by otutusaus on 2010-11-16 at 16:47 |
|
|
|
|
|
#5 |
|
"Richard B. Woods"
Aug 2002
Wisconsin USA
22×3×641 Posts |
It depends on whether or not there's already been a LL test done. If the factoring is being done preliminarily to a DC, its cost is balanced against just one LL test (plus a fraction for errors).
|
|
|
|
![]() |
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| P-1 factoring question | siegert81 | Math | 3 | 2014-03-09 12:38 |
| Factoring Question | Rde | Software | 12 | 2009-06-12 22:38 |
| question on P-1 factoring, stage 2 | nngs | Software | 1 | 2006-11-15 11:07 |
| Factoring Question | AntonVrba | Math | 7 | 2006-08-30 07:15 |
| factoring question | philmoore | Factoring | 8 | 2005-06-14 22:13 |