Polynomial Request Thread - Page 106

swellman · 2018-04-09, 19:43

Code:

R0: -198163287610408008474299544600510487866
R1: 713588604854929381
A0: 464860712074244018011487871108388695701300147178095
A1: 883598583001102527733435593339761342580914
A2: -2028992567682389466010511292053996
A3: -2093021702063113798100666
A4: 831165963415509
A5: 68400
skew 1743820123.59, size 2.088e-019, alpha -8.380, combined = 6.415e-015 rroots = 5

Max0526 · 2018-04-11, 21:05

Quote:

Originally Posted by swellman

Code:

R0: -198163287610408008474299544600510487866
R1: 713588604854929381
A0: 464860712074244018011487871108388695701300147178095
A1: 883598583001102527733435593339761342580914
A2: -2028992567682389466010511292053996
A3: -2093021702063113798100666
A4: 831165963415509
A5: 68400
skew 1743820123.59, size 2.088e-019, alpha -8.380, combined = 6.415e-015 rroots = 5

I tried to improve this poly --> no luck, sorry.

Max0526 · 2018-04-11, 21:14

I guess there is no need to post polys with the E below 7.5e-15, right? We have three so far that are above 7.5e-15.
Who is test-sieving for C197? Approximately when?
If you suddenly want to test-sieve more than three polys, please let us know. I have some 7.1s-7.4s that are not posted here.
Also, I can try spinning anything better than 6.7 into shape.

swellman · 2018-04-11, 22:43

Sorry, I had posted my highest score to date - a rather anemic 6.415. It was for pride’s sake. Won’t publish anything < 7.00 going forward.

Approaching 250k for c5. Slow going but better speed than I was getting previously. But to cover my reserved range up to 500k will take several more weeks. Planning to plow through unless someone suggests otherwise. I’m willing to complete it, I just ask you delay proposing this job to NFS@Home until I finish my range.

RichD · 2018-04-12, 00:06

@Sean
Save your cycles for another time. I believe we are at or near record levels for this size number.
Let's wrap this up, put a bow on it and call it a winner.

What's next on the chopping block? I thought I may be out of town so I didn't commit too much last time around, but I'm ready now.

RichD · 2018-04-12, 00:15

I guess the real question is, Can we shoe horn this into a 15e siever?
That may dictate whether we have a "good enough" poly.

VBCurtis · 2018-04-12, 01:10

There's no doubt we can do this with 15e/33. I think C200 can be done with nfs@home 15e, though the matrix will likely be bigger than if done with 16e.

GNFS-200 is the spot where solving the matrix is the bigger hurdle than getting the sieving done. Aliquot-276 at C205 can be done with 15e, but the matrix will be gross (and worse than 16). I'm willing to donate a thread-year (that is, two months on 6-core i7) to a GNFS200+ matrix for the sake of pushing our tools, but that won't be enough to finish such a beast.

So, unless the owner of new shiny iron wishes to offer his services for the C205, I think we should poly-select Max's C200 next. I am more interested in the C205 from Aliq-276, but until and unless someone who matters agrees to postprocess it it's folly.

My GPU work completes tonight at 1M; I've run CADO from 500k to 750k with no poly within 10% of the 6.9 I posted a week or so ago. Assuming my last runs with CADO and msieve-GPU produce nothing worth posting to Max, I'll test-sieve the three 7.5+ polys Thursday evening.

VBCurtis · 2018-04-12, 02:49

Dear Max:
Please polish. Msieve-generated.

Code:

# norm 3.102524e-19 alpha -6.943160 e 7.217e-15 rroots 3
skew: 152039729.27
c0: 2931130217185327116601317015565420773934927200
c1: -918193698486396137048328916808242683510
c2: 46768211205451464797643011244055
c3: 40873832936202201931322
c4: 1577994290944152
c5: 991728
Y0: -116079600316758606974465208257186926517
Y1: 13552554441652526533

VBCurtis · 2018-04-12, 07:03

Starting on the C205 from Aliquot 276 at 1M on msieve. Stage1norm of 1e30 still has 11 pieces available!

I don't mind switching to Max's C200 if there is consensus that nobody wants to solve the C205 matrix.

swellman · 2018-04-12, 11:22

Quote:

Originally Posted by RichD

@Sean
Save your cycles for another time. I believe we are at or near record levels for this size number.
Let's wrap this up, put a bow on it and call it a winner.

Agreed. Characterizing 1 to N in poly searching is less important here than “good enough to sieve”. I’ll stop my search later when I get home, followed by -npr. Just in case.

As to the next poly search, C205 seems a bridge too far with 15e. So we either search the C200, which will be a new 15e record GNFS job, or we ask Greg if he’s willing to take on the C205 with 16. Either way we go it’s a BIG job.

I vote for the C200. I will reserve the range c5 < 100,000 to start regardless of which composite we tackle.

Dubslow · 2018-04-12, 14:05

Forgetful moron here, what's the size of a GNFS-205 matrix? North or south of 32 GiB?

2018-04-11, 21:14	#1158
Max0526 "Max" Jun 2016 Toronto 19·47 Posts	C197 polys I guess there is no need to post polys with the E below 7.5e-15, right? We have three so far that are above 7.5e-15. Who is test-sieving for C197? Approximately when? If you suddenly want to test-sieve more than three polys, please let us know. I have some 7.1s-7.4s that are not posted here. Also, I can try spinning anything better than 6.7 into shape.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
GIMPS wiki account request thread	ixfd64	mersennewiki	169	2018-09-21 05:43
Polynomial Discriminant is n^k for an n-1 degree polynomial	carpetpool	Miscellaneous Math	14	2017-02-18 19:46
Lost Prime Raider password request thread	cheesehead	Forum Feedback	6	2009-07-28 13:02
Polynomial	R.D. Silverman	NFSNET Discussion	13	2005-09-16 20:07
Deutscher Thread (german thread)	TauCeti	NFSNET Discussion	0	2003-12-11 22:12

2018-04-11, 22:43	#1159
swellman Jun 2012 3091₁₀ Posts	Sorry, I had posted my highest score to date - a rather anemic 6.415. It was for pride’s sake. Won’t publish anything < 7.00 going forward. Approaching 250k for c5. Slow going but better speed than I was getting previously. But to cover my reserved range up to 500k will take several more weeks. Planning to plow through unless someone suggests otherwise. I’m willing to complete it, I just ask you delay proposing this job to NFS@Home until I finish my range.

2018-04-12, 00:06	#1160
RichD Sep 2008 Kansas 3,389 Posts	@Sean Save your cycles for another time. I believe we are at or near record levels for this size number. Let's wrap this up, put a bow on it and call it a winner. What's next on the chopping block? I thought I may be out of town so I didn't commit too much last time around, but I'm ready now.

2018-04-12, 00:15	#1161
RichD Sep 2008 Kansas 3,389 Posts	I guess the real question is, Can we shoe horn this into a 15e siever? That may dictate whether we have a "good enough" poly.

2018-04-12, 01:10	#1162
VBCurtis "Curtis" Feb 2005 Riverside, CA 11×443 Posts	There's no doubt we can do this with 15e/33. I think C200 can be done with nfs@home 15e, though the matrix will likely be bigger than if done with 16e. GNFS-200 is the spot where solving the matrix is the bigger hurdle than getting the sieving done. Aliquot-276 at C205 can be done with 15e, but the matrix will be gross (and worse than 16). I'm willing to donate a thread-year (that is, two months on 6-core i7) to a GNFS200+ matrix for the sake of pushing our tools, but that won't be enough to finish such a beast. So, unless the owner of new shiny iron wishes to offer his services for the C205, I think we should poly-select Max's C200 next. I am more interested in the C205 from Aliq-276, but until and unless someone who matters agrees to postprocess it it's folly. My GPU work completes tonight at 1M; I've run CADO from 500k to 750k with no poly within 10% of the 6.9 I posted a week or so ago. Assuming my last runs with CADO and msieve-GPU produce nothing worth posting to Max, I'll test-sieve the three 7.5+ polys Thursday evening.

2018-04-12, 07:03	#1164
VBCurtis "Curtis" Feb 2005 Riverside, CA 11×443 Posts	Starting on the C205 from Aliquot 276 at 1M on msieve. Stage1norm of 1e30 still has 11 pieces available! I don't mind switching to Max's C200 if there is consensus that nobody wants to solve the C205 matrix.

2018-04-12, 14:05	#1166
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 3×29×83 Posts	Forgetful moron here, what's the size of a GNFS-205 matrix? North or south of 32 GiB?