GNFS targets which need more ECM - Page 18

amphoria · 2016-12-12, 09:33

Quote:

Originally Posted by swellman

Amphoria - can you run some additional poly select? I can drop a request over in the poly select thread in the msieve project as well.

I can run poly select on a GTX-1080 for another 5 days.

amphoria · 2016-12-17, 16:32

Quote:

Originally Posted by amphoria

I can run poly select on a GTX-1080 for another 5 days.

After another 5 days I found this polynomial. I am sure that fivemack will be able to improve on this.

Code:

# norm 7.258266e-019 alpha -7.644954 e 1.229e-014 rroots 5
skew: 358825801.60
c0: -541034266273071643706370766895582192548127541360
c1: 13718046031139156486286095072652870602606
c2: -39816315113392860005980728236821
c3: -627498896636735434245193
c4: 59744206662587
c5: 489060
Y0: -20969440925509810850271371696232832741
Y1: 20158651314281929

Max0526 · 2016-12-20, 03:01

amphoria's first poly optimized by Msieve:

Code:

Msieve v. 1.52 (SVN 958)
R0: -17690115030839161531537758166981467963
R1: 378465978813130231
A0: -33355624080348901725207331898539288433015044480
A1: 3352384959730610289848710538986274755944
A2: -87910725208095177601320984748978
A3: -57464971932238955826111
A4: 1550716504180238
A5: 1144572
skew 226329347.85, size 6.013e-019, alpha -7.535, combined = 1.242e-014

fivemack · 2016-12-20, 14:14

After about 1.5 thread-years, I haven't got anything better than a single 1.292e-14, but I will leave the machines crunching for 72 more hours.

max0526: how long would it take you to do your magic improvement work if I send you a batch of 64 reasonable polynomials? There is no particular urgency to get this started by Christmas - there are still four unstarted jobs at https://escatter11.fullerton.edu/nfs/crunching_e.php and I have sieving for a C169 and a C171 queued up for the 64 cores to chew on - but it would seem worth running my candidates past you before I do the heavy trial sieving.

(I think this job is 15e 33-bit with 3 large primes on the A side, unless the statistics for newer polynomials are completely different than for amphoria's)

Max0526 · 2016-12-20, 17:11

Quote:

Originally Posted by fivemack

max0526: how long would it take you to do your magic improvement work if I send you a batch of 64 reasonable polynomials?

It will take an hour in Windows and ~4 hours in Linux for 64 polys. You'll have them back in a day. Send them in!

fivemack · 2016-12-20, 18:00

Quote:

Originally Posted by Max0526

It will take an hour in Windows and ~4 hours in Linux for 64 polys. You'll have them back in a day. Send them in!

OK, I'll send you a pile tonight and another pile Friday morning if the cluster finds anything better in the meantime.

fivemack · 2017-01-04, 22:43

C193_143_93 has been queued up using the best polynomial I found (2.5 thread-years of msieve, then quite careful trial sieving on 48 candidates).

The linear algebra is going to be not completely trivial, but it'll fit on hardware that I have.

amphoria · 2017-01-05, 11:52

Quote:

Originally Posted by fivemack

C193_143_93 has been queued up using the best polynomial I found (2.5 thread-years of msieve, then quite careful trial sieving on 48 candidates).

For interest would you mind posting the polynomial selected and associated e score here.

fivemack · 2017-01-05, 13:18

Code:

n: 1982879862067973474891514859868732513299540021583405085797833744255797525633234105710592159850731168994776922107643301701606141995707074821793831620255524328119596210394385668575040940696331187
skew: 382391084.95
lss: 0
c0: 318324081209204166613742905286715136916623438560
c1: 3615738780888862816019755064121406537188
c2: -25748356788550039031717676712618
c3: -107519724368142216637064
c4: -258026194215637
c5: 167076
Y0: -25994200629018740942376323273940528517
Y1: 894445074784661731
lpbr: 33
lpba: 33
mfbr: 66
mfba: 96
alambda: 3.6
rlambda: 2.6
alim: 400000000
rlim: 400000000

Code:

skew 382391084.95, size 6.103e-19, alpha -7.408, combined = 1.266e-14 rroots = 3

Average yield was 2.73 per Q, 0.819 seconds per relation

fivemack · 2017-01-07, 11:53

I will do some GPU-assisted ECM on this number, 23040 curves at B1=3e8 seems the right sort of amount.

fivemack · 2017-01-19, 23:55

3*48 hours has already got a 2.403e-14 for C187_142_67, so I'll leave it running for a week and see what comes out after 50x the search effort.

Code:

# norm 2.213343e-18 alpha -7.185978 e 2.403e-14 rroots 3
skew: 22191584.09
c0: 16995239059266192255429640770247824861912500
c1: -15862356931604487635103677383363656480
c2: 2225523048801700876285916592105
c3: 373753792949863429536619
c4: -2780023445302335
c5: 10137204
Y0: -931996517400320220400110552483980301
Y1: 2212333432420110271

2016-12-20, 14:14	#191
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 6419₁₀ Posts	After about 1.5 thread-years, I haven't got anything better than a single 1.292e-14, but I will leave the machines crunching for 72 more hours. max0526: how long would it take you to do your magic improvement work if I send you a batch of 64 reasonable polynomials? There is no particular urgency to get this started by Christmas - there are still four unstarted jobs at https://escatter11.fullerton.edu/nfs/crunching_e.php and I have sieving for a C169 and a C171 queued up for the 64 cores to chew on - but it would seem worth running my candidates past you before I do the heavy trial sieving. (I think this job is 15e 33-bit with 3 large primes on the A side, unless the statistics for newer polynomials are completely different than for amphoria's) Last fiddled with by fivemack on 2016-12-20 at 14:16

2017-01-07, 11:53	#197
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 7²×131 Posts	C194_144_91 I will do some GPU-assisted ECM on this number, 23040 curves at B1=3e8 seems the right sort of amount.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Ready GNFS targets	XYYXF	XYYXF Project	86	2020-03-07 16:23
SNFS targets which need more ECM	XYYXF	XYYXF Project	57	2017-07-04 19:15
Ready SNFS targets	XYYXF	XYYXF Project	25	2016-11-20 21:35
3,697+ (GNFS 220.9)	pinhodecarlos	NFS@Home	0	2014-12-24 19:13
3,766+ (GNFS 215.5)	pinhodecarlos	NFS@Home	34	2014-04-01 21:27

2017-01-04, 22:43	#194
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 7²×131 Posts	C193_143_93 has been queued up using the best polynomial I found (2.5 thread-years of msieve, then quite careful trial sieving on 48 candidates). The linear algebra is going to be not completely trivial, but it'll fit on hardware that I have.