Polynomial Request Thread

[swellman]

Originally Posted by EdH

This is not near your preferred range, but some testing gave me this:

SKEW 758184394.20
R0 -1877983994142100543416629807737490458501
R1 2280842250530422739
A0 -414349527140594965902462839198958465213716771630280
A1 -762666578956989362003714339018876350746674
A2 4980782355384860279547087572759473
A3 -2366329972187727464281370
A4 26988496136834196
A5 3483480
size 3.743e-20, alpha -8.255, combined = 2.160e-15 rroots = 3

I can't seem to get cownoise to respond to my attempts to use it, so I don't know if there is something odd or if this is really even poorer than it seems.

Thanks Ed! Cownoise kicked these values at me:

758183361.66069 2.15963261e-15

Not sure why it gave you problems. I’ve seen cownoise fail due a typo, i.e. the polynomial is incorrectly entered and ggnfs or msieve would crash if either attempted to use it for sieving or scoring. And sometimes cownoise is just down though it appears to be online. It’s all a mystery.

Suspect this composite will require a much larger c5 if using msieve to search - the skew for your poly is huge. Max could somehow use his origami skills to reduce the skew by a factor of 2-5 without too many ripples in the poly itself but he seems to have disappeared.

I started a CADO run at an ad search range of 10-20M (and I presume multiples of same). Using incr 4620 and plan on searching to upwards of ad = 300M. Will take a few weeks.

[EdH]

I'm trying to perfect my script for one of my GPUs and it includes sending the values to cownoise, but it hung on the wget call. I tried manually with copy/paste and it wouldn't come back with a value. I'll try again in a bit. Maybe it was just down for a while.

This GPU is only running during daytime, so I'm bouncing around with small ranges (.5M). I'll see if anything significant actually shows up and try to get the cownoise part to work.

[frmky]

Originally Posted by swellman

C203_147_104 is a GNFS from the XYYXF project.

A day with msieve on an A40 gave this as the best. Still a bit below your range.

Edit: The search went up to about 15M.

R0: -1808663928460726406224959814449203828778
R1: 12684447350737457147
A0: 79023423883195567180085827887402008548049142522005
A1: 776148312606369973035788015609125109945388
A2: 58445342249001142809762693848315
A3: -4407743860304266708109676
A4: -3810881822154592
A5: 4204200
skew 627433473.30, size 6.011e-20, alpha -8.183, combined = 3.020e-15 rroots = 3

[swellman]

Originally Posted by frmky

A day with msieve on an A40 gave this as the best. Still a bit below your range.

Edit: The search went up to about 15M.

Many thanks!

[swellman]

Originally Posted by swellman

I started a CADO run at an ad search range of 10-20M (and I presume multiples of same). Using incr 4620 and plan on searching to upwards of ad = 300M. Will take a few weeks.

First result:

n: 81371580958288180495643291649819154765300654909188614047433715271318289820697671548034835547680276517220926062058191594407670460404935983632115144113293325360273639747305204929704831648236717426500461987
skew: 100921134.695
c0: -2925565105834768097306215268990140135183594895267
c1: -192757317784014060762437489615819400874411
c2: 2121885762705819391780748581159883
c3: 4704355652596637064642031
c4: -282675407406732576
c5: 163783620
Y0: -1349239595873600982352222385081515131500
Y1: 1065160792884806717024011
# MurphyE (Bf=8.590e+09,Bg=4.295e+09,area=2.684e+16) = 4.025e-09
# f(x) = 163783620*x^5-282675407406732576*x^4+4704355652596637064642031*x^3+2121885762705819391780748581159883*x^2-192757317784014060762437489615819400874411*x-2925565105834768097306215268990140135183594895267
# g(x) = 1065160792884806717024011*x-1349239595873600982352222385081515131500

Cownoise says 152959771.67179 2.30721673e-15

I’m going run this job from 20M to 50M next, hoping for better results.

[frmky]

I put in another A40-week searching a significant fraction of a5 up to about 168M. Nothing new better than 2.66e-15 was found. The top-5 were

# norm 7.268308e-20 alpha -8.182906 e 3.020e-15 rroots 3
# norm 7.234239e-20 alpha -8.248137 e 3.014e-15 rroots 3
# norm 7.037883e-20 alpha -8.145686 e 2.954e-15 rroots 3
# norm 7.094437e-20 alpha -7.831841 e 2.939e-15 rroots 3
# norm 6.849223e-20 alpha -8.051469 e 2.907e-15 rroots 3

I also put in 10% of the default CADO search using factory c200 parameters covering up to 1M. The best I found, with a5 = 161760, was 2.67e-15.

[swellman]

Originally Posted by frmky

I put in another A40-week searching a significant fraction of a5 up to about 168M. Nothing new better than 2.66e-15 was found. The top-5 were

# norm 7.268308e-20 alpha -8.182906 e 3.020e-15 rroots 3
# norm 7.234239e-20 alpha -8.248137 e 3.014e-15 rroots 3
# norm 7.037883e-20 alpha -8.145686 e 2.954e-15 rroots 3
# norm 7.094437e-20 alpha -7.831841 e 2.939e-15 rroots 3
# norm 6.849223e-20 alpha -8.051469 e 2.907e-15 rroots 3

I also put in 10% of the default CADO search using factory c200 parameters covering up to 1M. The best I found, with a5 = 161760, was 2.67e-15.

Thanks Greg. I’ll look around the forum for some poly spin scripts, maybe something found to date can be improved.

I am currently running a CADO search for ad range of 20-50M with incr = 4260. Results later this week.

[EdH]

I just started a spin on the 2.03 3.02 poly in post #2236. I've lost my faith in my spin script, but we'll see what it may run up. . .

If you create a list of all the polys you'd like spun, I can run the whole list at once and see which turns up as best via the script.

[EdH]

Here's the spin for the 3.02 poly in post #2236:

n: 81371580958288180495643291649819154765300654909188614047433715271318289820697671548034835547680276517220926062058191594407670460404935983632115144113293325360273639747305204929704831648236717426500461987
Y0: -1808663928462715526664641064904980808090
Y1: 12684447350737457147
c0: -6514727271088895248342887593240599785160947950739
c1: 48355315275291388574381025275288644031904
c2: 1407645785876092822425094261371215
c3: -983457864716576474333548
c4: -7107304567770592
c5: 4204200
skew: 306670856.125
# lognorm 63.72, E 55.96, alpha -7.76 (proj -2.37), 5 real roots
# MurphyE(Bf=1.000e+07,Bg=5.000e+06,area=1.000e+16)=3.182e-15

Best poly cownoise values: 423230276.30644      3.18358329e-15

[frmky]

Here are the other 4 top candidates, all from the same a5 value:

# norm 7.234239e-20 alpha -8.248137 e 3.014e-15 rroots 3
skew: 616198965.11
c0: 29157359981793147315806256689518215512966706826725
c1: 645144013702620219355476242226142410589140
c2: -1435138678749702902320006415035237
c3: -5505291356839455779260780
c4: -1728307640466592
c5: 4204200
Y0: -1808663928459469743717865642937798876962
Y1: 12684447350737457147

# norm 7.037883e-20 alpha -8.145686 e 2.954e-15 rroots 3
skew: 595986230.11
c0: 5002502991594536986195694755881958863165611839845
c1: 567028014937595722374142967200075905521772
c2: -1835193556129348071443356645349845
c3: -5646495832448618833901164
c4: -1225935422118592
c5: 4204200
Y0: -1808663928459166603336622676953275068526
Y1: 12684447350737457147

# norm 7.094437e-20 alpha -7.831841 e 2.939e-15 rroots 3
skew: 553255529.04
c0: 50713112089425762997516261067909447724717028154960
c1: 553368726372231060153730618840746576051522
c2: 1341506608847148654029404890658415
c3: -1478294187722831799576924
c4: -6731478370023592
c5: 4204200
Y0: -1808663928462488746415445038748470253861
Y1: 12684447350737457147

# norm 6.849223e-20 alpha -8.051469 e 2.907e-15 rroots 3
skew: 631937674.13
c0: 70146209957443727139914103892428209801895590662680
c1: 760969950487591761534006190059581853221130
c2: -470579192691389195706077384136337
c3: -4922886138618734317064540
c4: -3018015464871592
c5: 4204200
Y0: -1808663928460247976484931880121063197797
Y1: 12684447350737457147

[EdH]

Thanks! I have them spinning up.