Reserved for MF - Sequence 4788

[em99010pepe]

10*p+1 at B1=1000000000, B2=205702371522480

[jrk]

Originally Posted by bchaffin

Including previous posts, I think that puts us at nearly 9500 curves at 40-43M, plus 800 more at 100M.

That's plenty.

Originally Posted by bchaffin

I've now found a couple polys, the best of which is
# norm 4.025205e-15 alpha -7.671371 e 2.488e-12 rroots 3
Is it about time to start gnfs?

Yes, let's use that poly. Please post it, as it's good enough for sieving.

Use the params from this post, and if you don't mind doing a little more work, sieve samples of specialq's in ranges of 1000 for each 2M, starting at 20M, and see how far you have to go to get about 28500 relations. The upside of doing this is so that we get a nice range to work in, without having to worry about going overboard with the sieving and making the filtering/LA harder.

[bchaffin]

Here's the poly:

# norm 4.025205e-15 alpha -7.671371 e 2.488e-12 rroots 3
skew: 42373826.68
c0: -58398068784506051024477192866875977892800
c1: -167100295795302960137099458278320
c2:  511004049252803790191305716
c3: -13968245979099022648
c4: -357816499243
c5:  996
Y0: -2379225273126981752554350978483
Y1:  158202809821186829

And here are the test counts:

   1524 20m-test.dat
   1655 22m-test.dat
   1585 24m-test.dat
   1588 26m-test.dat
   1598 28m-test.dat
   1382 30m-test.dat
   1376 32m-test.dat
   1406 34m-test.dat
   1292 36m-test.dat
   1210 38m-test.dat
   1422 40m-test.dat
   1209 42m-test.dat
   1273 44m-test.dat
   1280 46m-test.dat
   1075 48m-test.dat
   1129 50m-test.dat
    975 52m-test.dat
   1247 54m-test.dat
   1205 56m-test.dat
    947 58m-test.dat
   1007 60m-test.dat
   1400 62m-test.dat
    996 64m-test.dat
   1329 66m-test.dat
   1103 68m-test.dat
    879 70m-test.dat
   1102 72m-test.dat
   1008 74m-test.dat
    832 76m-test.dat
   1542 78m-test.dat

It reaches 28500 at about 62m.

[jrk]

Originally Posted by bchaffin

It reaches 28500 at about 62m.

Alright good, so round that up to the next region and the specialq range is 20M to 64M.

Sieving can begin whenever!

[jrk]

Originally Posted by jrk

A poly for the line 2618 c155:

Note to readers: bchaffin posted a better poly for this c155 in the ECM work on 4788:i2618 c155 thread, where sieving has already started.

10metreh: moved the ECM discussion from that thread into here, so that thread is now just a sieve thread

[bchaffin]

Update:

The c155 mentioned in the previous post fell with a 72/84 split. Three more terms came quickly, then a c152 at 2622, then nine more quick ones. The sequence picked up some factors of 5 for a few terms and grew quickly, but the 5 is now gone again and it has decreased the last 4 steps.

It's now at 2631 with a c166. I've run 904@1e6, 2350@3e6, and 4480@11e6.

I've got another project brewing so I'll pass the baton here for the c166.

Edit: I also ran ~9 hours of polyfind, which (oddly) found over 7000 polynomials. The best so far is:

# norm 3.113254e-16 alpha -6.987220 e 5.420e-13 rroots 5
skew: 104433618.57
c0: -3594443180787779467039152916634938000972403
c1:  77297502049290424735528622763755480
c2:  16355360267472884920695810030
c3: -114270217445156762108
c4: -1551899100643
c5:  444
Y0: -318454654779098929858653632108904
Y1:  200065134768558431

[jrk]

Originally Posted by bchaffin

Edit: I also ran ~9 hours of polyfind, which (oddly) found over 7000 polynomials. The best so far is:

# norm 3.113254e-16 alpha -6.987220 e 5.420e-13 rroots 5

I'll do some poly search as well. With a little more time on poly search, that score could improve by around ~15%. That would save a few days of sieving.

For ECM, say, 20000 curves @ B1=1e8?

[jrk]

Originally Posted by jrk

I'll do some poly search as well. With a little more time on poly search, that score could improve by around ~15%. That would save a few days of sieving.

I'm trial-sieving some polys now. Will come back to post a full poly late tonight or in the morning.

[Andi_HB]

Sequence is now at a C146 in iteration 2635
1000 curves @1e6 are done
curves with 3e6 are running

[jrk]

+260 curves @ B1=8e7

Also doing polynomial selection. I will post one a bit later.

[Andi_HB]

2500 @ 3e6 done - no factor