mersenneforum.org  

Go Back   mersenneforum.org > Other Stuff > Archived Projects > NFSNET Discussion

 
 
Thread Tools
Old 2005-07-20, 01:05   #1
R.D. Silverman
 
R.D. Silverman's Avatar
 
Nov 2003

746010 Posts
Default P751

Can Richard or Paul reveal the parameters used for 2,751+?

Sizes of factor bases, large prime bounds, sieve region boundaries/length
of the sieve lines etc.??
R.D. Silverman is offline  
Old 2005-07-20, 13:45   #2
dleclair
 
dleclair's Avatar
 
Mar 2003

7×11 Posts
Default

Hi Bob,

I can help with this. We used f(x)=2*x^6+1 and g(x)=x-m with common root 2^125.

We used a factorbase bound of 30M on each side and large prime limit of 500M on each side.

Line lengths were 180M, 126M and 90M, longer lines being used at smaller b's but the bulk of the sieving with 90M. I don't know the exact boundaries at which the line lengths changed (Richard has the details).

We've sieved about 45M lines so far and have collected about 43.6M relations. Richard indicated that after some filtering tests, we may need to sieve a bit longer before we can get a manageable matrix.

-Don
dleclair is offline  
Old 2005-07-20, 15:12   #3
R.D. Silverman
 
R.D. Silverman's Avatar
 
Nov 2003

22·5·373 Posts
Default

Quote:
Originally Posted by dleclair
Hi Bob,

I can help with this. We used f(x)=2*x^6+1 and g(x)=x-m with common root 2^125.

We used a factorbase bound of 30M on each side and large prime limit of 500M on each side.

Line lengths were 180M, 126M and 90M, longer lines being used at smaller b's but the bulk of the sieving with 90M. I don't know the exact boundaries at which the line lengths changed (Richard has the details).

We've sieved about 45M lines so far and have collected about 43.6M relations. Richard indicated that after some filtering tests, we may need to sieve a bit longer before we can get a manageable matrix.

-Don

Hi,

Your line lengths *appear* to be sub-optimal, but since your factor base
is smaller than optimal, this may be OK. The lines close to the origin
seem about right, but at larger b's (say b > 10M) they should only
be 20-30M in length...


An optimal factor base would be "about" 35% bigger .


I understand very well the need to make the matrix as small as possible.

Note that for 2,791+ I am using factor base bounds of 24M, with a number
that is 70 bits smaller than yours....I estimate my matrix will have 4M rows.
My LP bound is 700M.

Thanks for the info.

Bob
R.D. Silverman is offline  
 

Thread Tools


All times are UTC. The time now is 14:06.


Mon Oct 25 14:06:46 UTC 2021 up 94 days, 8:35, 0 users, load averages: 1.15, 1.34, 1.35

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.