Going over 100% during sieving

wombatman · 2013-07-10, 04:01

I'm currently working on a 128 digit composite, and I'm in the sieving phase. The number of relations is now at 155% of the expected total needed. The polynomial gave an e-score that was right below the expected window. What parameters can I tinker with to try and reduce my sieving time? For a C128, it's not a huge deal, but with bigger numbers, having to add more relations could definitely mean a lot more time.

firejuggler · 2013-07-10, 23:29

155%? i never got that much over-sieving. At most 115-120. There might be something wrong with your relation.Did you try to force it LA mode?

wombatman · 2013-07-11, 00:14

Well, it eventually got whatever it wanted and did the linear algebra and quadratic steps without any issue. This was with the SVN 900 version of msieve using Brian Gladman's python script.

prgamma10 · 2013-07-11, 06:07

Maybe all ranges have been sieved twice?
(If the number of used cores is changed in-between the factorization, the savefile will be corrupted).

wombatman · 2013-07-11, 15:41

That's possible. I'm pretty sure I played around with the number of cores/threads used to do the sieving. So maybe that is what caused it.

2013-07-10, 04:01	#1
wombatman I moo ablest echo power! May 2013 29×61 Posts	Going over 100% during sieving I'm currently working on a 128 digit composite, and I'm in the sieving phase. The number of relations is now at 155% of the expected total needed. The polynomial gave an e-score that was right below the expected window. What parameters can I tinker with to try and reduce my sieving time? For a C128, it's not a huge deal, but with bigger numbers, having to add more relations could definitely mean a lot more time.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
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
Sieving	robert44444uk	Sierpinski/Riesel Base 5	8	2005-04-02 22:30

2013-07-10, 23:29	#2
firejuggler Apr 2010 Over the rainbow 2³·5²·13 Posts	155%? i never got that much over-sieving. At most 115-120. There might be something wrong with your relation.Did you try to force it LA mode?

2013-07-11, 00:14	#3
wombatman I moo ablest echo power! May 2013 29×61 Posts	Well, it eventually got whatever it wanted and did the linear algebra and quadratic steps without any issue. This was with the SVN 900 version of msieve using Brian Gladman's python script.

2013-07-11, 06:07	#4
prgamma10 Jan 2013 109 Posts	Maybe all ranges have been sieved twice? (If the number of used cores is changed in-between the factorization, the savefile will be corrupted).

2013-07-11, 15:41	#5
wombatman I moo ablest echo power! May 2013 29·61 Posts	That's possible. I'm pretty sure I played around with the number of cores/threads used to do the sieving. So maybe that is what caused it.