|
2009-03-16, 09:43
|
#2 | |
|
Bamboozled!
"𒉺𒌌𒇷𒆷đ’€"
May 2003
Down not across
10,753 Posts |
Quote:
It will almost certainly take a significant amount of personal effort to learn how to perform the factorization. Paul |
|
|
|
|
2009-03-16, 11:57
|
#3 | ||
|
Jan 2009
616 Posts |
this is new log:
Quote:
Quote:
Last fiddled with by wsgtrsys on 2009-03-16 at 11:59 |
||
|
|
|
|
2009-03-16, 14:05
|
#4 |
|
(loop (#_fork))
Feb 2006
Cambridge, England
72×131 Posts |
That log indicates that you are using entirely wrong parameters; the sooner you stop, the less computer time you will have wasted.
In particular, you will have done vastly less polynomial searching than optimal, and you are using absurdly small factor-base and large-prime bounds, meaning that your yield per special-Q is so small that you will not get a usable number of relations before you run out of usable Q. If you change the lines at the bottom of your .poly file to something like lpbr: 29 lpba: 29 mfbr: 58 mfba: 58 alim: 40000000 rlim: 40000000 and start again, using factMsieve rather than factLat because ggnfs filtering tools aren't really capable enough for the 512-bit level, then I would expect you might be able to complete within six to twelve CPU-months. If you don't understand any of the terms I've used here, please ask; but I assure you that any more time spent running with your current parameters will be as wasted as all the time you've put in so far has been: you've been trying to bale out the Titanic using a ladle with holes in. |
|
|
|
|
2009-03-16, 23:00
|
#6 | ||
|
Jan 2009
2×3 Posts |
ggnfs .poly file is:
Quote:
Quote:
Last fiddled with by wsgtrsys on 2009-03-16 at 23:01 |
||
|
|
|
|
2009-03-17, 02:44
|
#7 |
|
Tribal Bullet
Oct 2004
67258 Posts |
Don't bother with the line sieve in msieve; this polynomial has a very large amount of skew, and no amount of line sieving is going to find enough relations. You need to use the lattice sieve from GGNFS to have any hope of completing the job.
Perhaps you should instead start with something small, like RSA100? |
|
|
|
|
2009-03-17, 06:16
|
#8 | |
"Serge"
Mar 2008
Phi(4,2^7658614+1)/2
36·13 Posts |
Quote:
A ladder: Try a RSA-100. Then try a GNFS-120, that will take you 10 times longer. Then try a GNFS-140, that will take you extra 10 times longer still. Then try a GNFS-154 (RSA-512), it will take you 5 times longer still than the previous. For every 20-21 extra digits you will need 10x more time to spend. And you need the correct parameters and an increasing understanding of what you are doing. |
|
|
|
|
|
2009-03-17, 06:54
|
#9 |
Nov 2008
2×33×43 Posts |
And msieve 1.38 is now out of date.
|
|
|
|
|
2009-05-03, 18:59
|
#10 | |
Mar 2009
Indiana, United Stat
24·3 Posts |
Some suggestions
I thought I read on Wikipedia that there’s a European team that’s been factoring RSA-768 for months. In late March or early April, I contacted the webmaster of RSA asking if he knew anything about this or how I could join. He/she didn’t email me back information.
Quote:
From the RSA Factoring Challenge, Jens Franke seems to be a step ahead of everybody: http://en.wikipedia.org/wiki/RSA_Factoring_Challenge I typed “Jens Franke” into the internet and I found his info and email: http://www.hausdorff-center.uni-bonn.de/old/faculty/Franke/ I didn’t get around to emailing him though. However, I would email asking him a couple questions including about the computer time if I were you. Ask him if he knows anything about the European team that’s factoring RSA-768. I know enough about RSA-170 that I shouldn’t try to factor it unless I either get about 7 computers or several more people to use their computing time to help me. I have just one computer. It’s going to possibly be years until there are faster computers or I get enough people that want to help me with RSA-170. Some of the suggestions that I’ve gotten are that I should factor a Cunningham Tables number or an aliquot number. My advice to you would be to start with a smaller number like one with 99 to 121 digits just to be sure that you can use the GGNFS factoring software correctly. Jeff Gilchrist’s Guide should help: http://gilchrist.ca/jeff/factoring/nfs_beginners_guide.html Then look at the polynomial search for 109!+1: http://www.mersenneforum.org/showthread.php?t=11454 and how the sieving for it is going: http://www.mersenneforum.org/showthread.php?t=11529 . I try to remember to check back at these 109!+1 pages every two days or so. You said you have 7 PCs. Maybe you should start where I left off at RSA-170. Or maybe you should use your computer time to help the European team with RSA-768. That’s about all I know that I can help you with. Remember that the people here on the forum are here to help you if you have any more questions. |
|
|
|
|
|
2009-05-03, 19:18
|
#11 | |
|
Nov 2008
2·33·43 Posts |
Quote:
There are many people on the forum who are MUCH more experienced with NFS than you. You are behaving as if you want to become a moderator. Sander and Tom are the moderators in this forum because they are experienced and know a lot about how factoring works. |
|
|
|
|
 |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| newb needs help - factoring RSA512 error | barthmaul | Msieve | 22 | 2014-12-04 08:25 |
| today ,i factoring rsa512 is ok,only 3 day | jackchina | Factoring | 7 | 2012-04-13 17:57 |
| NFS Factoring Time | Stargate38 | Factoring | 6 | 2011-12-08 21:42 |
| LLT, double-checking and factoring at the same time | T.Rex | Math | 12 | 2006-02-08 20:35 |
| per iteration time for Trial Factoring? | MavsFan | Software | 3 | 2003-12-12 03:42 |
