3,748+ c204 Smaller-but-Needed

[swellman]

Bump. Any interest in revisiting this poly search and possible team sieve? Summary of poly search targets:

Global best C205 e-score: 2.215e-15

Local best (EdH): 2.60680364e-15

Global best C204 e-score: 2.630e-15

Threshold (msieve): 2.78e-15

Objective (msieve): > 3.20e-15

[VBCurtis]

Sure- taking 3e7 to 6e7 on CADO with incr=4620, P=8000000,nq=15625.

Interested parties can run with P changed to much higher or much lower, with little to no overlap with my search. One might try P=4M or P=14M for new runs; the closer one chooses P to my run, the more overlap in polys found there will be- 2x (or half) should have near-zero overlap.

[EdH]

interesting - I did a quick run and got a poly that cownoise wouldn't comment on, as well. It, too, had a large skew and much smaller original MurphyE than the earlier one posted that cownoise didn't like. BTW, no manual copy/paste error, as I use a script to run cownoise. There is still copy/paste, but it's done by the script.

[swellman]

I will attempt 6e7 to 7e7 with P = 4M. May also attempt P values of 8M and 14M if my hardware can handle that search space in a reasonable time.

[VBCurtis]

Originally Posted by VBCurtis

Sure- taking 3e7 to 6e7 on CADO with incr=4620, P=8000000,nq=15625.

Here's the poly:

n: 533439167600904850230361756102700151678687933392166847323827307497363839257031077774321424872955045754669625$
skew: 98578127.386
c0: -543850183462425327198930487409925109829785629120
c1: 68212999208925720003012825859916399522882
c2: 638443675326649192046822053722117
c3: -17982199392452538405419168
c4: -39487971248308278
c5: 171157140
Y0: -1563731901211757910450408924988302711541
Y1: 5408299121980104968049211
# MurphyE (Bf=1.718e+10,Bg=8.590e+09,area=1.074e+16) = 1.503e-08

cownoise says score 2.29e-15. No help.
This took about 40hr on 40 threads.

[EdH]

I had a 2.3 I didn't post because it was less than the earlier VBCurtis one. Should I post close ones?

[swellman]

I’m still working on my range but I’m playing around a bit in a vain attempt to tune/optimize CADO poly searching a c204. Learning something if nothing else, but collecting data as I go.

So far nothing found worth reporting, though the e-score of the poly keeps increasing.

[VBCurtis]

Originally Posted by EdH

I had a 2.3 I didn't post because it was less than the earlier VBCurtis one. Should I post close ones?

Max's 'spin' can yield a 5-8% improvement for one poly vs another, so I would say posting anything within 10% of the expected best is worthwhile. For this job, we should get to 2.5x, so I'd say 2.3's are worth posting. Max may reappear and spin a dozen polys for us to turn up the best.

[EdH]

Here's the 2.3 poly:

n: 533439167600904850230361756102700151678687933392166847323827307497363839257031077774321424872955045754669625577486179222154434651598903112919949771321416511589029559325246084363632977829645558547714072241
skew: 10042040.26
c0: -627354527254627652475332655176329358596636356
c1: 213018845490759638721666401827674851417
c2: 95937374251117784023748647685013
c3: -5623556931527657096482154
c4: -479102663461196184
c5: -2511784800
Y0: -1385041351159270154396558379928364326380
Y1: 262954605447105862796669
# MurphyE (Bf=8.590e+09,Bg=4.295e+09,area=2.684e+16) = 4.040e-09
# f(x) = -2511784800*x^5-479102663461196184*x^4-5623556931527657096482154*x^3+95937374251117784023748647685013*x^2+213018845490759638721666401827674851417*x-627354527254627652475332655176329358596636356
# g(x) = 262954605447105862796669*x-1385041351159270154396558379928364326380

cownoise says:

16158683.33848 2.31283325e-15

But, this one showed up from somewhere, as though it was from the last run I made. I would have sworn it had shown an 8.8e-16, but now it's showing much better:

n: 533439167600904850230361756102700151678687933392166847323827307497363839257031077774321424872955045754669625577486179222154434651598903112919949771321416511589029559325246084363632977829645558547714072241
skew: 61173787.626
c0: -84551986041964239497940179882870301230255594832
c1: -1390964012076111490370063297140812459596
c2: 522002006121443981512032398747760
c3: 1762901709293407365122057
c4: -15943093281758022
c5: -318686760
Y0: -1380919702784425269323122024964712327251
Y1: 237213586863942644893691
# MurphyE (Bf=8.590e+09,Bg=4.295e+09,area=2.684e+16) = 4.390e-09
# f(x) = -318686760*x^5-15943093281758022*x^4+1762901709293407365122057*x^3+522002006121443981512032398747760*x^2-1390964012076111490370063297140812459596*x-84551986041964239497940179882870301230255594832
# g(x) = 237213586863942644893691*x-1380919702784425269323122024964712327251

cownoise says:

71610307.84090 2.60680364e-15

[EdH]

Originally Posted by VBCurtis

Max's 'spin' can yield a 5-8% improvement for one poly vs another, so I would say posting anything within 10% of the expected best is worthwhile. For this job, we should get to 2.5x, so I'd say 2.3's are worth posting. Max may reappear and spin a dozen polys for us to turn up the best.

What is done to "spin" a polynomial? Is it something I can learn and/or implement with my archaic machinery and limited knowledge?

[EdH]

Here's another 2.4:

n: 533439167600904850230361756102700151678687933392166847323827307497363839257031077774321424872955045754669625577486179222154434651598903112919949771321416511589029559325246084363632977829645558547714072241
skew: 7740451.622
c0: 87401746937311294254606780488329283829954120
c1: 490423710004317779309293496266025547493
c2: -15973994938279417938365011227398
c3: -5798512478553491282564621
c4: 1272023397153994830
c5: -3220711200
Y0: -1378005267963093178565993313863772054930
Y1: 140867852901938662382179
# MurphyE (Bf=8.590e+09,Bg=4.295e+09,area=2.684e+16) = 4.170e-09
# f(x) = -3220711200*x^5+1272023397153994830*x^4-5798512478553491282564621*x^3-15973994938279417938365011227398*x^2+490423710004317779309293496266025547493*x+87401746937311294254606780488329283829954120
# g(x) = 140867852901938662382179*x-1378005267963093178565993313863772054930

cownoise says:

12314038.28594 2.42609295e-15