mersenneforum.org  

Go Back   mersenneforum.org > Factoring Projects > Factoring

Reply
 
Thread Tools
Old 2008-10-14, 08:22   #1
fivemack
(loop (#_fork))
 
fivemack's Avatar
 
Feb 2006
Cambridge, England

13×491 Posts
Default 3221^73-1 sieving

Polynomial file is below

Code:
n: 3762417971112081220095196299015453579204743076262721175557307866835982313083323950322956026874886179059958104863089358770552434166984658821437838082376922335605302026116301206397054337279339461381076889146727871533242827472031146921363109621830525548633
c6: 3221
c0: -1
Y1: -1
Y0: 1247063872385372731496678575404593456513041
type: snfs
skew: 0.26
rlambda: 2.6
alambda: 2.6
alim: 75000000
rlim: 75000000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
Please sieve both algebraic and rational sides with gnfs-lasieve4I15e. On either side, 1MQ will take one 2.4GHz 64-bit core for about a week; the whole job will take about four core-years. The city for uploads is gothenburg.

14/11/2008: sieving is going pretty well, probably not worth making more reservations above 102 or below 37. We're nearly half-way to the desired number of relations in exactly one month

17/11/2008: after a bit more consideration of the quite high duplication rate, let's make that range 30-110

09/12/2008: the processing starts tonight

Code:
fivemack  36-44 (done)
fivemack  44-51 (done) 
fivemack  51-52 (rat) (done)
andi47     51-52 (alg)
bsquared   52-55 (done)
fivemack   55-65 (done)
batalov    65-70 (done)
bsquared   70-75 (done)
fivemack   75-78 (done)
FactorEyes 78-79 (done)
andi47     79-81, algebraic side only (done)
fivemack   79-81, rational side only (done)
FactorEyes 81-100 (done)
ZetaFlux   100-102 (done)
fivemack   102-110 (done)
Caught pancakes:
Code:
R 36-110
A 36-110

26/10/2008 23:45  11362879 relations
28/10/2008 21:18  33029988 relations (thanks mostly to bsquared)
14/11/2008 20:13  95238553 relations, 84650728 unique
22/11/2008 15:31  
 R-side    73195362 relations,  68944435 unique
 A-side    57017031 relations,  52702109 unique
 together 131431009 relations, 110045276 unique
01/12/2008 21:45
 R-side    98765072 relations,  90352435 unique
 A-side    86319328 relations,  77991055 unique
 together 185084400 relations, 149546079 unique
  weight of 16235147 cycles is about 1136477721 (70.00/cycle)

Last fiddled with by fivemack on 2008-12-09 at 17:20 Reason: the pancakes are stacking quite high now
fivemack is offline   Reply With Quote
Old 2008-10-14, 10:24   #2
Andi47
 
Andi47's Avatar
 
Oct 2004
Austria

9B216 Posts
Default

Quote:
Originally Posted by fivemack View Post
Polynomial file is below


Please sieve both algebraic and rational sides with gnfs-lasieve4I15e. On either side, 1MQ will take one 2.4GHz core for about a week; the whole job will take about four core-years. The city for uploads is gothenburg.

Code:
fivemack   75-78
Questions:

1.) How much RAM does this sieving job take?
2.) How many relations will we need approximately?
3.) You take 75-78 on both sides?

Last fiddled with by Andi47 on 2008-10-14 at 10:27 Reason: added 3rd question
Andi47 is offline   Reply With Quote
Old 2008-10-14, 10:51   #3
fivemack
(loop (#_fork))
 
fivemack's Avatar
 
Feb 2006
Cambridge, England

13·491 Posts
Default

1) on Linux, 653MB VSZ and 283MB RSZ for each job
2) about two hundred million
3) yes

Last fiddled with by fivemack on 2008-10-14 at 10:52
fivemack is offline   Reply With Quote
Old 2008-10-14, 11:40   #4
Andi47
 
Andi47's Avatar
 
Oct 2004
Austria

2×17×73 Posts
Default

reserving 79-81 on algebraic side

Edit: I think this should take me approx. 3 to 4 weeks.

Last fiddled with by Andi47 on 2008-10-14 at 11:45
Andi47 is offline   Reply With Quote
Old 2008-10-14, 13:29   #5
ValerieVonck
 
ValerieVonck's Avatar
 
Mar 2004
Belgium

292 Posts
Default

Hi,

If I would reserve a range, do I have to adapt both numbers or how should I
set a range??

Thank you.
Regards
Cedric

alim: 75000000
rlim: 75000000
ValerieVonck is offline   Reply With Quote
Old 2008-10-14, 13:49   #6
Andi47
 
Andi47's Avatar
 
Oct 2004
Austria

1001101100102 Posts
Default Benchmark

5000 Q take approx. 2 hours on a Core2Duo at 1800 MHz, so 2MQ take approx. 33 days. (1 thread, 32 bit windows). The job uses ~330MB Ram.

Last fiddled with by Andi47 on 2008-10-14 at 13:57 Reason: added RAM useage
Andi47 is offline   Reply With Quote
Old 2008-10-14, 13:55   #7
Andi47
 
Andi47's Avatar
 
Oct 2004
Austria

46628 Posts
Default

Quote:
Originally Posted by CedricVonck View Post
Hi,

If I would reserve a range, do I have to adapt both numbers or how should I
set a range??

Thank you.
Regards
Cedric

alim: 75000000
rlim: 75000000
Just copy/paste the polynomial file posted by fivemack into a file, for example named 3221_73-.poly

then type into your command line:

Code:
gnfs-lasieve4I15e -o <outputfile> -a 3221_73-.poly -f <start of your range> -c <length of your range>
the -a statement will sieve on the algebraic side. If you want to sieve on the rational side, type -r instead of -a.

(for my range of 79-81 algebraic, I typed gnfs-lasieve4I15e -o 3221_73_79.81 -a 3221_73-.poly -f 79000000 -c 2000000)

P.S.: You don't need to change rlim and alim as long as you don't sieve below the rlim and alim values.

Last fiddled with by Andi47 on 2008-10-14 at 13:58
Andi47 is offline   Reply With Quote
Old 2008-10-14, 14:22   #8
FactorEyes
 
FactorEyes's Avatar
 
Oct 2006
vomit_frame_pointer

16816 Posts
Default Range convention?

Is there some reason we're skipping 78-79?

EDIT: I'll take 78-79, as well as 81-100, both sides.

Last fiddled with by FactorEyes on 2008-10-14 at 15:03
FactorEyes is offline   Reply With Quote
Old 2008-10-14, 14:27   #9
bsquared
 
bsquared's Avatar
 
"Ben"
Feb 2007

41·83 Posts
Default

I'll take 70-75 A+R
bsquared is offline   Reply With Quote
Old 2008-10-14, 15:44   #10
Zeta-Flux
 
Zeta-Flux's Avatar
 
May 2003

7×13×17 Posts
Default

Quote:
Originally Posted by Andi47 View Post
Just copy/paste the polynomial file posted by fivemack into a file, for example named 3221_73-.poly

then type into your command line:

Code:
gnfs-lasieve4I15e -o <outputfile> -a 3221_73-.poly -f <start of your range> -c <length of your range>
the -a statement will sieve on the algebraic side. If you want to sieve on the rational side, type -r instead of -a.

(for my range of 79-81 algebraic, I typed gnfs-lasieve4I15e -o 3221_73_79.81 -a 3221_73-.poly -f 79000000 -c 2000000)

P.S.: You don't need to change rlim and alim as long as you don't sieve below the rlim and alim values.
If you want to do sieve both the rational and algebraic, do you do those separately then?

Also, is there a good website for downloading a precompiled version of gnfs for Windows XP?
Zeta-Flux is offline   Reply With Quote
Old 2008-10-14, 15:55   #11
bsquared
 
bsquared's Avatar
 
"Ben"
Feb 2007

340310 Posts
Default

Quote:
Originally Posted by Zeta-Flux View Post
If you want to do sieve both the rational and algebraic, do you do those separately then?
Yep. Make sure <outputfile> is different if run on the same computer and directory... files will be clobbered otherwise.

Quote:
Originally Posted by Zeta-Flux View Post
Also, is there a good website for downloading a precompiled version of gnfs for Windows XP?

This might work for you:
http://www.mersenneforum.org/showpos...97&postcount=8

Last fiddled with by bsquared on 2008-10-14 at 15:58 Reason: output files note...
bsquared is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Going over 100% during sieving wombatman Msieve 4 2013-07-11 15:41
NFS sieving? Dubslow Factoring 8 2012-09-28 06:47
Line sieving vs. lattice sieving JHansen NFSNET Discussion 9 2010-06-09 19:25
10^420 + 1 sieving juno1369 Factoring 20 2010-04-28 01:11
Sieving OmbooHankvald Prime Sierpinski Project 4 2005-06-30 07:51

All times are UTC. The time now is 07:18.

Mon Apr 12 07:18:38 UTC 2021 up 4 days, 1:59, 1 user, load averages: 2.68, 2.74, 2.32

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.