mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Aliquot Sequences (https://www.mersenneforum.org/forumdisplay.php?f=90)
-   -   Reserved for MF - Sequence 4788 (https://www.mersenneforum.org/showthread.php?t=11615)

jrk 2009-10-13 17:08

Here's a preliminary poly found with Jason's new msieve-gpu branch.

[code]# norm 2.523819e-15 alpha -7.877120 e 1.839e-12
skew: 2785154.52
c0: -138570565461588730821881652221421126805
c1: -356118891405782364779335898545631
c2: 157627527368055796683392529
c3: 368273140050413578367
c4: -24904624740140
c5: 949200
Y0: -1035248474311472807522765292042
Y1: 832063911771386519
[/code]

Note that msieve's final_norm cutoff for this composite is 1.835e-12.

bsquared 2009-10-13 17:17

My curvefits suggest > 1.95e-12 would be good.

10metreh 2009-10-13 17:30

[QUOTE=bsquared;192729]My curvefits suggest > 1.95e-12 would be good.[/QUOTE]

The previous c157 had a 2.00e-12 poly.

bsquared 2009-10-13 17:45

[quote=10metreh;192730]The previous c157 had a 2.00e-12 poly.[/quote]

And it was good :wink:.

Greebley 2009-10-13 18:36

[quote=Greebley;192712]I was thinking it would be neat to get 314718 to index 9000, however it is at 8902 now so that would be a long 92 indices more.
It is index 2540 for 4788.[/quote]

[quote]Both those figures are wrong: 314718 is at index 8908, and 4788 is at index 2448.[/quote]

Yeah the index is 8908 which is 92 indices from 9000 so I mis-typed that one.

However, the other value (2540) is the current index + 92 i.e. we could stop at index 2540 for that is when we reach 9000 for 314718. So for that one it just wasn't clear what I was trying to say.

To summarize:
My suggestion is to run the sequence 4788 to index 2540 if we can.

10metreh 2009-10-13 18:43

[QUOTE=Greebley;192736]My suggestion is to run the sequence 4788 to index 2540 if we can.[/QUOTE]

Unless we acquire the downdriver at something ridiculous like 172 digits, in which case we just [B]have[/B] to continue. :smile:

axn 2009-10-14 08:21

I'll run pol51 for 0-200M using following parameters:
[CODE]pol51m0b -b c157_1 -p 7 -n 2.8e23 -a 0 -A 200000
pol51opt -b c157_1 -n 3.6e22 -N 3.6e19 -e 1.7e-12
[/CODE]

bsquared 2009-10-14 13:15

I left the range running overnight, and woke up to these:

[code]
polysel_0-2500.cand:BEGIN POLY #skewness 979863.43 norm 8.85e+20 alpha -5.12 Murphy_E 2.14e-12
polysel_0-2500.cand:BEGIN POLY #skewness 1249264.69 norm 1.13e+21 alpha -5.19 Murphy_E 2.07e-12
polysel_0-2500.cand:BEGIN POLY #skewness 1239920.67 norm 1.08e+21 alpha -5.17 Murphy_E 2.07e-12
polysel_0-2500.cand:BEGIN POLY #skewness 1064109.97 norm 9.82e+20 alpha -4.76 Murphy_E 1.97e-12
polysel_0-2500.cand:BEGIN POLY #skewness 979626.75 norm 9.01e+20 alpha -4.60 Murphy_E 1.94e-12
[/code]

I'll let it continue to run... in the mean time, anyone care to test sieve the top 2 or 3? We can probably duplicate the siever and parameters from the last C157.

details:
[code]
n: 1128681916333165583281832717120127496638972860118923964844791402387721331170448754199775637678274646951935918929237315749341487244542351920306209843985409563
#skewness 979863.43 norm 8.85e+20 alpha -5.12 Murphy_E 2.14e-12
c5: 156600
c4: 5511109690356
c3: 2556240664351703353
c2: -2709038327031089573784479
c1: 772591600803924819123938074991
c0: -127641604647397600699119628378829341
Y1: 64393121809011209
Y0: -1484416691862834512911536697536
skew: 197324.88
rlim: 45000000
alim: 45000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6

n: 1128681916333165583281832717120127496638972860118923964844791402387721331170448754199775637678274646951935918929237315749341487244542351920306209843985409563
#skewness 1249264.69 norm 1.13e+21 alpha -5.19 Murphy_E 2.07e-12
c5: 156600
c4: 5314399732356
c3: -2883058196108786471
c2: -2573465483932011297872477
c1: 2271343202342143863498206577119
c0: 23472793401947121132547837819239017
Y1: 64393121809011209
Y0: -1484416708040060932502186689770
skew: 197324.88
rlim: 45000000
alim: 45000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6

n: 1128681916333165583281832717120127496638972860118923964844791402387721331170448754199775637678274646951935918929237315749341487244542351920306209843985409563
#skewness 1239920.67 norm 1.08e+21 alpha -5.17 Murphy_E 2.07e-12
c5: 156600
c4: 5335915006356
c3: -2297759499328129799
c2: -2787019491787729028389067
c1: 2123826836917266179771394505903
c0: 18333066845778520513050090022694295
Y1: 64393121809011209
Y0: -1484416706270666731434176688868
skew: 197324.88
rlim: 45000000
alim: 45000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6
[/code]

I just realized all of those (and a half dozen more 1.9x 's) have the same leading coefficient! Never seen that before...

henryzz 2009-10-14 16:30

once i have finally managed to compile msieve-gpu i will have a CUDA gpu i can run on polynomial selection
it is proving difficult to get environment variables and such like correct

all you need is a nvidia series >=8
those look good polys

bsquared 2009-10-14 16:42

[quote=bsquared;192797]I left the range running overnight, and woke up to these:

[code]
polysel_0-2500.cand:BEGIN POLY #skewness 979863.43 norm 8.85e+20 alpha -5.12 Murphy_E 2.14e-12
polysel_0-2500.cand:BEGIN POLY #skewness 1249264.69 norm 1.13e+21 alpha -5.19 Murphy_E 2.07e-12
polysel_0-2500.cand:BEGIN POLY #skewness 1239920.67 norm 1.08e+21 alpha -5.17 Murphy_E 2.07e-12
polysel_0-2500.cand:BEGIN POLY #skewness 1064109.97 norm 9.82e+20 alpha -4.76 Murphy_E 1.97e-12
polysel_0-2500.cand:BEGIN POLY #skewness 979626.75 norm 9.01e+20 alpha -4.60 Murphy_E 1.94e-12
[/code]

[/quote]

I decided to do some test sieving on the top 3 as well as jrk's msieve poly. I was dissapointed by the performance of the top poly (about 1100 rels in a range of 1000 Q starting at 10M, 0.170 sec/rel), and the others were even worse in comparison. So on a whim I drug out a 1.92e-12 scoring poly with a better alpha (-6.44), and it beat the pants off of the 2.14e-12 poly. Rough curve fits of test sieving 1k blocks from 5M to 50M in steps of 5M suggest we'll need about 43MQ and about 7 million cpu seconds (based on data from a 3.16 GHz core 2) or 81 CPU days.

I think there are better polys to be found, but this is the baseline so far. Lesson to be learned: Murphy_E is a crapshoot!

bsquared 2009-10-14 16:55

1 Attachment(s)
Data attached, if anyone is interested.

henryzz 2009-10-15 06:17

[quote=henryzz;192816]once i have finally managed to compile msieve-gpu i will have a CUDA gpu i can run on polynomial selection
it is proving difficult to get environment variables and such like correct

all you need is a nvidia series >=8
those look good polys[/quote]
i finally got it working
i will start polynomial selection with it

axn 2009-10-15 09:35

[QUOTE=bsquared;192817]So on a whim I drug out a 1.92e-12 scoring poly with a better alpha (-6.44), and it beat the pants off of the 2.14e-12 poly. <snip>
I think there are better polys to be found, but this is the baseline so far. Lesson to be learned: Murphy_E is a crapshoot![/QUOTE]

May I offer the following three polys for your consideration:
[CODE]BEGIN POLY #skewness 492944.87 norm 2.47e+021 alpha -6.03 Murphy_E 2.00e-012
X5 22783680
X4 3049889691192
X3 25115676975823063898
X2 -1057549120220075124773221
X1 -4015168516778629810185220755126
X0 -68074828915752550906211652474462780
Y1 56460946476157681
Y0 -548263756929827444068662381961
M 130374183394572597071227248002684258680753480045928203593454005872997714178038084759490414258234615905198731655011653462453390100735529354844596548096485215
END POLY
BEGIN POLY #skewness 591796.65 norm 3.01e+021 alpha -6.32 Murphy_E 2.02e-012
X5 22783680
X4 2209855409592
X3 25038106255076701466
X2 -1612254100171446541836739
X1 -7285061836793671322852863008672
X0 120437822537708031145046391210398448
Y1 56460946476157681
Y0 -548263757346170463383849121655
M 130374183394572597071227248002684258680753480045928203593454005872997714178038084759490414258234615905198731655011653462453390100735529354844596548096492589
END POLY
BEGIN POLY #skewness 594103.05 norm 3.05e+021 alpha -6.34 Murphy_E 2.03e-012
X5 22783680
X4 2558901387192
X3 25067329196727393818
X2 -1381972795712101042561075
X1 -7294236258178618545696273107894
X0 -164061303268787107813228616719980278
Y1 56460946476157681
Y0 -548263757173174123380901987071
M 130374183394572597071227248002684258680753480045928203593454005872997714178038084759490414258234615905198731655011653462453390100735529354844596548096489525
END POLY
[/CODE]

axn 2009-10-15 10:42

[QUOTE=bsquared;192797]details:
[code]
#skewness [B]979863.43[/B] norm 8.85e+20 alpha -5.12 Murphy_E 2.14e-12
c5: 156600
c4: 5511109690356
skew: [B]197324.88[/B]

#skewness [B]1249264.69[/B] norm 1.13e+21 alpha -5.19 Murphy_E 2.07e-12
c5: 156600
c4: 5314399732356
skew: [B]197324.88[/B]

#skewness [B]1239920.67[/B] norm 1.08e+21 alpha -5.17 Murphy_E 2.07e-12
c5: 156600
c4: 5335915006356
skew: [B]197324.88[/B]
[/code]
[/QUOTE]

Saaay... Did you just test all those goodly polynomials with wrong skew??? :glare:

EDIT:- Yes you did For the 2.14:
Wrong skew {total yield: 1050, q=30001003 (0.33863 sec/rel)}
Proper skew {total yield: 1174, q=30001003 (0.29744 sec/rel)}

bsquared 2009-10-15 13:08

[quote=axn;192905]Saaay... Did you just test all those goodly polynomials with wrong skew??? :glare:

EDIT:- Yes you did For the 2.14:
Wrong skew {total yield: 1050, q=30001003 (0.33863 sec/rel)}
Proper skew {total yield: 1174, q=30001003 (0.29744 sec/rel)}[/quote]

Err, yes, yes I did :doh!:. Copied it over from the last C157...

I get:

Wrong skew {total yield: 1116, q=10001009 (0.17081 sec/rel)}
Proper skew {total yield: 1341, q=10001009 (0.14201 sec/rel)}

FWIW, the 1.92e-12 poly was also tested wrong:

Wrong skew {total yield: 1396, q=10001009 (0.14012 sec/rel)}
Proper skew {total yield: 1421, q=10001009 (0.13689 sec/rel)}

And here are your 3 candidates (as you posted them, top to bottom):

total yield: 1078, q=10001009 (0.13871 sec/rel)
total yield: 951, q=10001009 (0.14025 sec/rel)
total yield: 1211, q=10001009 (0.14852 sec/rel)

And my two 2.07e-12 polys:

total yield: 1195, q=10001009 (0.14133 sec/rel)
total yield: 1038, q=10001009 (0.14251 sec/rel)

And finally, jrk's msieve poly:
total yield: 1163, q=10001009 (0.14720 sec/rel)

I'll take the 2.14 poly, 1.92 poly, and your fastest poly and do some more trial sieving.

bsquared 2009-10-15 14:15

1 Attachment(s)
[quote=bsquared;192919]
I'll take the 2.14 poly, 1.92 poly, and your fastest poly and do some more trial sieving.[/quote]

A winner is emerging: axn's 2.00e-12 poly appears to be consistently faster, with a higher yield, than the others I tested.

[code]
n: 1128681916333165583281832717120127496638972860118923964844791402387721331170448754199775637678274646951935918929237315749341487244542351920306209843985409563
#BEGIN POLY #skewness 492944.87 norm 2.47e+021 alpha -6.03 Murphy_E 2.00e-012
c5: 22783680
c4: 3049889691192
c3: 25115676975823063898
c2: -1057549120220075124773221
c1: -4015168516778629810185220755126
c0: -68074828915752550906211652474462780
Y1: 56460946476157681
Y0: -548263756929827444068662381961
skew: 492944.87
rlim: 45000000
alim: 45000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6
[/code]

jrk 2009-10-15 15:54

Here are the three best polys I found overnight:

[code]# norm 2.999613e-15 alpha -7.053170 e 2.102e-12
skew: 4947197.07
c0: 254361561800804638444888132008691713600
c1: 277567242655343801120663754660120
c2: -41169227890365677403210934
c3: -33865259805947197259
c4: 5813164911518
c5: 214800
Y0: -1393498408963548236214652730759
Y1: 581416348390623701
# norm 2.797246e-15 alpha -7.634500 e 2.004e-12
skew: 7829822.03
c0: -392472115828519587544012662072667953792
c1: 1135538850431039541173288486836512
c2: 215718619935831247136208602
c3: -66702913924475596555
c4: 4019834079518
c5: 214800
Y0: -1393499379793961455729608702127
Y1: 581416348390623701
# norm 2.691749e-15 alpha -7.287346 e 1.958e-12
skew: 7493033.63
c0: -1191940646279096538567449411970147312640
c1: 245517647645424865764025430284152
c2: -67767123827532957236661598
c3: -26904687145857425547
c4: 6126231615518
c5: 214800
Y0: -1393498239483008345741406384063
Y1: 581416348390623701
[/code]

And look at the alpha on this one!
[code]# norm 2.644805e-15 alpha -8.500494 e 1.914e-12
skew: 18761259.11
c0: 70384025424543567250040749496664182714880
c1: 34739194678189905179867695959241928
c2: 369205076515588254645800050
c3: -294264765478046013769
c4: 659280547786
c5: 214200
Y0: -1394280990497229313048382187377
Y1: 593607538600963639
[/code]

bsquared 2009-10-15 16:06

[quote=jrk;192929]Here are the three best polys I found overnight:
[/quote]

I'll do a comparison.

10metreh 2009-10-15 17:46

I always hate it when no-one mentions the siever used for testing, or bothers to test between sievers. I presume you have been using 14e for testing, but it would be good to give 15e a try, just in case.

bsquared 2009-10-15 17:47

The clear winner is now this one from jrk:

[code]
n: 1128681916333165583281832717120127496638972860118923964844791402387721331170448754199775637678274646951935918929237315749341487244542351920306209843985409563
# norm 2.797246e-15 alpha -7.634500 e 2.004e-12
skew: 7829822.03
c0: -392472115828519587544012662072667953792
c1: 1135538850431039541173288486836512
c2: 215718619935831247136208602
c3: -66702913924475596555
c4: 4019834079518
c5: 214800
Y0: -1393499379793961455729608702127
Y1: 581416348390623701
rlim: 45000000
alim: 45000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6
[/code]

I didn't test the last one with really low alpha.

Anyone still have any poly selection going on? Henryzz?

The above poly would need ~ 39MQ to get ~50M relations, and take a little less than 6Msec. We've chopped a million seconds of sieving time off by finding better polys (and using them correctly)... pretty nice!

[edit]
Using 14e siever. sorry I thought that was clear. I don't think we need 15e for this one... Feel free to test it if you want.

henryzz 2009-10-15 18:49

i stopped my poly selection run
i suspect my gpu is rather pathetic
also it seg faults every time i hibernate which makes it almost unusable for that size number

jrk 2009-10-16 04:54

Thanks for doing the tests, bsquared.

[quote]Anyone still have any poly selection going on?[/quote]
I am going to run another overnight search.

As long as we are using Jason's msieve-gpu branch, and there is not yet a gpu version of lasieve4, then I think it does not take away from optimality to do a bit more poly searching than usual.

jrk 2009-10-17 22:01

[QUOTE=jrk;192977]I am going to run another overnight search.[/QUOTE]
I have not found anything better, and I'm quitting the search.

bsquared 2009-10-17 23:39

[quote=bsquared;192943]The clear winner is now this one from jrk:

[code]
n: 1128681916333165583281832717120127496638972860118923964844791402387721331170448754199775637678274646951935918929237315749341487244542351920306209843985409563
# norm 2.797246e-15 alpha -7.634500 e 2.004e-12
skew: 7829822.03
c0: -392472115828519587544012662072667953792
c1: 1135538850431039541173288486836512
c2: 215718619935831247136208602
c3: -66702913924475596555
c4: 4019834079518
c5: 214800
Y0: -1393499379793961455729608702127
Y1: 581416348390623701
rlim: 45000000
alim: 45000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6
[/code]
[/quote]


I've started sieving using this poly, 14e siever.

5M-11M reserved.

jrk 2009-11-11 17:46

[QUOTE=bsquared;195494][CODE]
prp50 factor: 64585367109663600884438834766883466988940788491041
prp107 factor: 17475814829955286940449820582268098487907223506432350199456831446049454150141950051204743013474395998182843


[/CODE]

Now on a C151.[/QUOTE]

I've started a run of 1000 curves @ B1=43000000, B2=388112953420. ETA 15 hours.

bsquared 2009-11-11 18:00

400 curves at B1=1M done.
Starting 2k curves at B1=3M (ETA 4hrs).

How does one report work to the DB now? There used to be a 'report work' button that I can't find anymore.

Dougal 2009-11-11 18:03

[B]How does one report work to the DB now? There used to be a 'report work' button that I can't find anymore.[/B]


go into the sequence,and beside the quick ecm button there is a >> button,click it,and at the bottom of the page you cant report the curves you've done.

bsquared 2009-11-11 18:17

[quote=Dougal;195503]go into the sequence,and beside the quick ecm button there is a >> button,click it,and at the bottom of the page you cant report the curves you've done.[/quote]

Looks like you have to be logged in, as well. Thanks.

unconnected 2009-11-12 07:18

After 2200 curves @ 11e6 I found a nice p47 :smile:
[QUOTE]Factor=12020207027570199228977626027667941677740462849 Method=ECM B1=11000000 Sigma=1642054642[/QUOTE]

Batalov 2009-11-12 08:53

I am even luckier :smile:
[code]Input number is 685715905099648994219451416570299620797010724632513288765357600633214758490518989168523251153670344481394131623 (111 digits)
Using B1=8000000, B2=23420528050, polynomial Dickson(12), sigma=4094310456
Step 1 took 23013ms
Step 2 took 13485ms
********** Factor found in step 2: 4122529854091491338824929291779507173076398066909
Found probable prime factor of 49 digits: 4122529854091491338824929291779507173076398066909
Probable prime cofactor 166333763336873313888188992301635414394739143771207611922744147 has 63 digits
[/code]
Then, a few easy lines, and...

10metreh 2009-11-12 15:25

[QUOTE=Batalov;195600]Then, a few easy lines, and...[/QUOTE]

A :censored: C161. :rant:

fivemack 2009-11-12 15:26

Running 2000@1e8, should be able to say something on Monday

(other than 'at least gmp-ecm builds on OS X' and 'Mac Pro is fast')

Andi47 2009-11-12 15:27

[QUOTE=10metreh;195619]A :censored: C161. :rant:[/QUOTE]

This seems to be near the crossover between the 14e and 15e sievers....

Andi47 2009-11-13 10:14

pp41-factor of c161
 
[CODE]GMP-ECM 6.2.3 [powered by GMP 4.3.0] [ECM]
Input number is 41989040709803320172325437685122012188238989866495646783169593798391300698922442158422655234035792573779700795308983586334745780088868313272292796749799498704069 (161 digits)

...

Run [B][COLOR="Red"]3[/COLOR][/B] out of 2000:
Using B1=43000000, B2=240490660426, polynomial Dickson(12), sigma=2746202462
Step 1 took 855859ms
Step 2 took 182687ms
********** Factor found in step 2: 17929080031745140528449157941308074577509
Found probable prime factor of 41 digits: 17929080031745140528449157941308074577509
Probable prime cofactor 2341951769720349963952569043284852276019038780750832785187060203064697177590116136874365861798306068919218595003497391841 has 121 digits[/CODE]

fivemack 2009-11-13 11:45

I had completed 700@1e8 without finding that, which is actually quite unlucky (ecm -v says 644 curves at 1e8 for a p45)

Killed them and started again on the C145 from 2459.

Greebley 2009-11-13 15:16

Boy that factor of 3 just won't go away. I guess at this many digits the chance of at least one factor + 1 being divisible by 3 is high.

mdettweiler 2009-11-13 16:07

[quote=Greebley;195718]Boy that factor of 3 just won't go away. I guess at this many digits the chance of at least one factor + 1 being divisible by 3 is high.[/quote]
Hmm...if I understand this correctly, then it's going to lose the 3 on the next line! :grin: Here's the current partial factorization of line 2459:

[URL="http://www.factordb.com/search.php?id=81655010"][SIZE=1][COLOR=#002099]1389005994...[/COLOR][/SIZE][/URL][SIZE=1]<168> = [/SIZE][URL="http://www.factordb.com/search.php?id=16882"][SIZE=1][COLOR=#000000]24[/COLOR][/SIZE][/URL][SIZE=1] · [/SIZE][URL="http://www.factordb.com/search.php?id=1"][SIZE=1][COLOR=#000000]3[/COLOR][/SIZE][/URL][SIZE=1] · [/SIZE][URL="http://www.factordb.com/search.php?id=13"][SIZE=1][COLOR=red]11[/COLOR][/SIZE][/URL][SIZE=1] · [/SIZE][URL="http://www.factordb.com/search.php?id=324"][SIZE=1][COLOR=#000000]593[/COLOR][/SIZE][/URL][SIZE=1] · [/SIZE][URL="http://www.factordb.com/search.php?id=595"][SIZE=1][COLOR=#000000]811[/COLOR][/SIZE][/URL][SIZE=1] · [/SIZE][URL="http://www.factordb.com/search.php?id=81655015"][SIZE=1][COLOR=#000000]81266036243201[/COLOR][/SIZE][/URL][SIZE=1]<14> · [/SIZE][URL="http://www.factordb.com/search.php?id=81655016"][SIZE=1][COLOR=#002099]6731092709...[/COLOR][/SIZE][/URL][SIZE=1]<145>[/SIZE]

Notice the 11: 11+1=12, which is divisible by 3.

fivemack 2009-11-13 16:35

(mdettweiler: you need to have *none* of the factors 2 mod 3 to lose the 3)

[code]
(job 5 out of 16) Run 12 out of 125:
Using B1=100000000, B2=776268975310, polynomial Dickson(30), sigma=1209849063
Step 1 took 1004841ms
Step 2 took 277472ms
********** Factor found in step 2: 1235256026055614152023140734345351300023869925971
Found probable prime factor of 49 digits: 1235256026055614152023140734345351300023869925971
Probable prime cofactor 5449147842682529938683571528241234609478304810765641311200653978116824057163714349422254281192109 has 97 digits
[/code]

Next line has C145, will give it the same treatment

Edit: found a P30 (ten times) and P37 (once) in 16x125 @ 1e6

axn 2009-11-13 16:45

[QUOTE=mdettweiler;195726]Hmm...if I understand this correctly, then it's going to lose the 3 on the next line! :grin: Here's the current partial factorization of line 2459:
<snip>
Notice the 11: 11+1=12, which is divisible by 3.[/QUOTE]

Which is why it can NOT lose the 3 on that iteration :sad:

bsquared 2009-11-13 19:11

I see we've now hit 170 digits... what is the record and if not this one, what is the highest sequence?

mdettweiler 2009-11-13 19:38

[quote=fivemack;195731](mdettweiler: you need to have *none* of the factors 2 mod 3 to lose the 3)[/quote]
[quote=axn;195732]Which is why it can NOT lose the 3 on that iteration :sad:[/quote]
Ah, I see. Thanks for the explanation. :smile:

10metreh 2009-11-13 20:07

[QUOTE=bsquared;195744]I see we've now hit 170 digits... what is the record and if not this one, what is the highest sequence?[/QUOTE]

162126 (171 digits)

henryzz 2009-11-13 20:31

[quote=10metreh;195753]162126 (171 digits)[/quote]
looking at the database it looks like it is actually 173 digits
looking at the karsten's records page it looks like 175 digits

kar_bon 2009-11-13 21:45

[QUOTE=henryzz;195758]looking at the karsten's records page it looks like 175 digits[/QUOTE]

corrected to 173 digits!

Raman 2009-11-15 06:20

How long are you going to extend the sequence 314718? Till iteration 9000? Is it getting difficult? Why don't we extend the sequence 1578 at iteration 7261? Is someone working on that number already?

10metreh 2009-11-15 12:33

[QUOTE=Raman;195917]How long are you going to extend the sequence 314718? Till iteration 9000? Is it getting difficult? Why don't we extend the sequence 1578 at iteration 7261? Is someone working on that number already?[/QUOTE]

1578 is under the control of a "driver" (2 * 3), which is a set of factors that persist and always push the sequence upwards. To escape 2 * 3, a line has to factor as 2 * 3^2 * p, where p is a prime of the form 4n+1. 4788 is not under the control of a driver, and although it is going up, there is a chance it will drop the 3 from 2^4 * 3, and then it will have a chance of acquiring the downdriver (2 * <something that isn't 3> * ....), after which it will go down. This means it has a much higher chance of terminating in the near future.

fivemack 2009-11-15 15:39

C145 from 4788:2466 polynomial
 
4000 @ 4e7 done on the C145 from 4788:2466; time for GNFS.

I've been running a polynomial selection in parallel; won't be able to do the sieving myself, but a pretty good polynomial is

[code]
n: 2731277733959968169313770677931211293039092240425293759926078826637186378568780767241107236657396164773599806169440569332028429227856163376999057
type: gnfs
skew: 2806664.12
c0: -537701751961052562535970119801728000
c1: 1744523195224888821560862375080
c2: 1338728504837455624301882
c3: -317279220701420893
c4: -224866274934
c5: 18480
Y0: -10812700210350163142223819967
Y1: 14144226355213709
lpbr: 28
lpba: 28
mfbr: 56
mfba: 56
alambda: 2.6
rlambda: 2.6
alim: 20000000
rlim: 20000000
[/code]

Siever 14e, sieve A side 10M-20M and see how much further is needed.

bsquared 2009-11-15 16:30

[quote=fivemack;195947]Siever 14e, sieve A side 10M-20M and see how much further is needed.[/quote]

Ok, thanks!

I'll take the whole job.

bsquared 2009-11-18 03:49

[code]prp68 factor: 45038238154540495578229005737573886297206834547447537402461361228041
prp77 factor: 60643529717749771496825937363787963901083989172307055129277826960716417883977
[/code]

+ a couple more iterations. currently on a C146, for which 3k curves at 3e6 will be done in a couple hours.

10metreh 2009-11-18 07:23

[QUOTE=bsquared;196256][code]prp68 factor: 45038238154540495578229005737573886297206834547447537402461361228041
prp77 factor: 60643529717749771496825937363787963901083989172307055129277826960716417883977
[/code]

+ a couple more iterations. currently on a C146, for which 3k curves at 3e6 will be done in a couple hours.[/QUOTE]

+ something that could be important in the future: the 3 is squared.

Prime95 2009-11-18 15:13

[QUOTE=10metreh;196265]+ something that could be important in the future: the 3 is squared.[/QUOTE]

Pardon my newbie-like question: How will 3^2 affect the driver?

10metreh 2009-11-18 15:37

[QUOTE=Prime95;196279]Pardon my newbie-like question: How will 3^2 affect the driver?[/QUOTE]

1. 2^4 * 3 is not a driver. It is not even a guide. 2^4 is the guide.
2. When the 3 is squared, it plays no part in proceedings as far as drivers and guides, in particular powers of 2, are concerned. This makes it easier for drivers such as 2 * 3 and 2^3 * 3 to be acquired.

Greebley 2009-11-18 18:18

ya, sigma(3^2) is odd, as compare to 3 itself with sigma of 4 (so 2 powers of 2).

It means for example that with the square if you get a single prime equal 1 mod 4 the power of 2 can go to 2^1. If that prime wasn't equal to 2 mod 3 (i.e. 1 mod 3) you would get the down driver. If it is 2 mod 3 then you get 2*3 which is not good because the 3 term sticks around.

With 3^1, the best you can get is 2^3 so one can't get the downdriver until the 3 term is lost or squared.

jrk 2009-11-20 05:30

I started a poly search for the c146. In about 24 hours I'll post a result.

Andi47 2009-11-20 06:28

[QUOTE=jrk;196479]I started a poly search for the c146. In about 24 hours I'll post a result.[/QUOTE]

How much ECM did this one have? I guess we should do at least t45 (7500@11e6) before doing GNFS?

schickel 2009-11-20 07:15

[QUOTE=Andi47;196480]How much ECM did this one have? I guess we should do at least t45 (7500@11e6) before doing GNFS?[/QUOTE]
I did some higher level curves. I'm not sure what the conversion ration is....

Andi47 2009-11-20 08:10

[QUOTE=schickel;196482]I did some higher level curves. I'm not sure what the conversion ration is....[/QUOTE]

According to my calculations based on the curves which have been reported to the DB, t45 is ~57% done.

jrk 2009-11-20 22:32

[QUOTE=Andi47;196480]How much ECM did this one have? I guess we should do at least t45 (7500@11e6) before doing GNFS?[/QUOTE]
Since we can search for polys on a GPU now but not ECM or sieving, it doesn't hurt to do poly searching early in parallel.

My best poly so far is 9.198e-12. I will do some testing to find the right parameters and post it later.

jrk 2009-11-21 01:52

This is the best from today but I will do another day of searching.

[code]n: 34128802065277683324940497060114456028128085387534241677541889565653711918779403288707008461679255695847408643283266931540024693124361473714544539
# norm 3.792523e-14 alpha -7.832101 e 9.198e-12
skew: 3154674.50
c0: 6160701223310457747430798454115232525
c1: -5353471580850156751287280021324
c2: -5703088353790837679214303
c3: 44415752942820154
c4: 1079694510848
c5: 53760
Y0: -14472016831258149683092208646
Y1: 27360769963746067
lpbr: 28
lpba: 28
mfbr: 56
mfba: 56
alambda: 2.6
rlambda: 2.6
alim: 20000000
rlim: 20000000
[/code]

jrk 2009-11-22 02:06

I had a 9.589e-12 poly but it test-sieved slower than the 9.198e-12 one.

henryzz 2009-11-22 18:41

[quote=jrk;196618]I had a 9.589e-12 poly but it test-sieved slower than the 9.198e-12 one.[/quote]
probably something to do with the 9.198e-12 having quite a good alpha
from memory someone said a while back that higher alpha means a higher chance of smaller factors and lower norms means the numbers to be factored are smaller
the e value attempts to get the mix right but doesnt always

10metreh 2009-11-22 19:44

[QUOTE=henryzz;196654]probably something to do with the 9.198e-12 having quite a good alpha
from memory someone said a while back that higher alpha means a higher chance of smaller factors and lower norms means the numbers to be factored are smaller
the e value attempts to get the mix right but doesnt always[/QUOTE]

With degree 4, the poly with the highest e value is very often suboptimal in fact.

henryzz 2009-11-22 20:27

[quote=10metreh;196660]With degree 4, the poly with the highest e value is very often suboptimal in fact.[/quote]
how do you find the optimal polynomial then?
look for the highest alpha or the smallest norms?

Andi47 2009-11-23 07:50

Stop GNFSing alq4788.2469...
 
...I found this factor by ECM:

[CODE]Run 226 out of 2000:
Using B1=43000000, B2=240490660426, polynomial Dickson(12), sigma=1163832276
Step 1 took 774079ms
Step 2 took 165421ms
********** Factor found in step 2: 1590362231581241185334390845152913218796531
Found probable prime factor of 43 digits: 1590362231581241185334390845152913218796531
Probable prime cofactor 21459766452919733123897504849837291094001756254327858522371153821180836119029805033757220337331923926969 has 104 digits
[/CODE]

and...

[code]2470. 116139550403454963217308225651624963448867441989096972645425812805257941827791127060931251649177260779207291173182882743273752673953845861192804201869628446785478569115056 = 2^4 * 3^2 * 7 * 115217807939935479382250223860739051040543097211405726830779576195692402606935641925527035366247282519054852354348097959596976859081196290865877184394472665461784294757
2471. 255322662394897022311066496075397737105843503420475090657007540849654364176969382506967910371603978062225552817235385078466900719723930980558783840618151426663313997184736 = 2^5 * 311 * 5867 * 1736734882092577 * c148[/code]


The 3 is gone!! :party:

jrk 2009-11-23 08:20

[code]Using B1=43000000, B2=388112953420, polynomial Dickson(30), sigma=1033163144
Step 1 took 138254ms
Step 2 took 77893ms
********** Factor found in step 2: 314401460320696602393783174941
Found probable prime factor of 30 digits: 314401460320696602393783174941
Probable prime cofactor 8008382854566027043204017961675695251441834519040939952694150641178698834997735974059700130106717640670945550858659647 has 118 digits
[/code]

Andi47 2009-11-23 08:29

[QUOTE=jrk;196723][code]Using B1=43000000, B2=388112953420, polynomial Dickson(30), sigma=1033163144
Step 1 took 138254ms
Step 2 took 77893ms
********** Factor found in step 2: 314401460320696602393783174941
Found probable prime factor of 30 digits: 314401460320696602393783174941
Probable prime cofactor 8008382854566027043204017961675695251441834519040939952694150641178698834997735974059700130106717640670945550858659647 has 118 digits
[/code][/QUOTE]

Me too, but with B1=1e6:

[code]Run 55 out of 950:
Using B1=1000000, B2=1045563762, polynomial Dickson(6), sigma=1225682356
Step 1 took 17734ms
Step 2 took 7328ms
********** Factor found in step 2: 314401460320696602393783174941
Found probable prime factor of 30 digits: 314401460320696602393783174941
Probable prime cofactor 8008382854566027043204017961675695251441834519040939952694150641178698834997735974059700130106717640670945550858659647 has 118 digits
[/code]

schickel 2009-11-23 10:38

[QUOTE=Andi47;196720]...I found this factor by ECM:[code]Found probable prime factor of 43 digits:[/code]
[/QUOTE]Wow! Congrats on that factor! :bow: :bow:

10metreh 2009-11-23 16:27

[QUOTE=henryzz;196662]how do you find the optimal polynomial then?
look for the highest alpha or the smallest norms?[/QUOTE]

You can't really except by test sieving, but luckily degree-4 numbers don't take too long.

10metreh 2009-11-23 18:06

BTW: 171 digits is a new record height for stability to be achieved. The old record was 164 in 1134 (which also holds the records for highest escapes from 2^5 * 3^2 * 7 and 2^5 * 3).

jrk 2009-11-23 21:44

About 1/3 of t45 done on the c142 so far. In case no factor is found, here is a decent poly:

EDIT: I'm still poly-searching, so a better poly may turn up later.

[code]n: 2348305321333187128721643453051615911877612201981824607595341595725431644675331447201502622476698707484077353247893830745261323782457787822233
# norm 1.016167e-13 alpha -7.694154 e 1.640e-11
skew: 1981323.38
c0: -290064811140012999156642070938904104
c1: 619521716085998538893959511396
c2: -301271769660356610615830
c3: -758842805286610009
c4: -135175380936
c5: 76140
Y0: -1985312530063977010532771629
Y1: 7024691883228631
lpbr: 28
lpba: 28
mfbr: 56
mfba: 56
alambda: 2.6
rlambda: 2.6
alim: 12500000
rlim: 12500000
[/code]

When ECM is finished, sieve with 13e from Q=2M to 20M.

axn 2009-11-23 22:38

[QUOTE=henryzz;196662]how do you find the optimal polynomial then?
look for the highest alpha or the smallest norms?[/QUOTE]

Why bother? Degree-4 jobs are so fast that you could pick _any_ one of them (found by msieve) and you should be fine.

jrk 2009-11-24 05:15

About 1/2 of t45 is done on the c142.

10metreh 2009-11-24 07:24

[QUOTE=jrk;196804]When ECM is finished, sieve with 13e from Q=2M to 20M.[/QUOTE]

Have you tested against 14e?

jrk 2009-11-24 08:57

[QUOTE=10metreh;196855]Have you tested against 14e?[/QUOTE]
Yes

[code]$ for i in `seq 2000000 2000000 20000000`; do ~/ggnfs/bin/gnfs-lasieve4I13e -a test.poly -f $i -c 1000; done
Warning: lowering FB_bound to 1999999.
total yield: 1132, q=2001037 (0.02852 sec/rel)
Warning: lowering FB_bound to 3999999.
total yield: 2029, q=4001003 (0.02795 sec/rel)
Warning: lowering FB_bound to 5999999.
total yield: 1257, q=6001013 (0.02944 sec/rel)
Warning: lowering FB_bound to 7999999.
total yield: 1814, q=8001029 (0.03114 sec/rel)
Warning: lowering FB_bound to 9999999.
total yield: 1255, q=10001009 (0.03446 sec/rel)
Warning: lowering FB_bound to 11999999.
total yield: 1352, q=12001001 (0.03404 sec/rel)
total yield: 1213, q=14001007 (0.03523 sec/rel)
total yield: 1345, q=16001081 (0.03771 sec/rel)
total yield: 1546, q=18001031 (0.03830 sec/rel)
total yield: 1268, q=20001001 (0.04148 sec/rel)[/code]

[code]$ for i in `seq 4000000 2000000 12000000`; do ~/ggnfs/bin/gnfs-lasieve4I14e -a test.poly -f $i -c 1000; done
Warning: lowering FB_bound to 3999999.
total yield: 3986, q=4001003 (0.03817 sec/rel)
Warning: lowering FB_bound to 5999999.
total yield: 2489, q=6001013 (0.03885 sec/rel)
Warning: lowering FB_bound to 7999999.
total yield: 3824, q=8001029 (0.03783 sec/rel)
Warning: lowering FB_bound to 9999999.
total yield: 2675, q=10001009 (0.04045 sec/rel)
Warning: lowering FB_bound to 11999999.
total yield: 2877, q=12001001 (0.03917 sec/rel)[/code]

14e is about 85% as fast as 13e. It would lower the duplication rate, though, so maybe it would be worth the hit. I took a guess that 13e is better.

jrk 2009-11-24 20:56

I stopped poly searching. The best one is still the one posted earlier. I think this number is ready for sieving.

Andi47 2009-11-25 08:27

[QUOTE=jrk;196861]
14e is about 85% as fast as 13e. It would lower the duplication rate, though, so maybe it would be worth the hit. I took a guess that 13e is better.[/QUOTE]

So - do we use 13e or 14e?

Andi47 2009-11-25 10:20

[QUOTE=Andi47;196955]So - do we use 13e or 14e?[/QUOTE]

p-1: B1=5e8, B2=5e13, no factor
p+1: 3 runs at B1=1e8, B2=6e12, no factor.

Running some more ECM until it is decided, whether we use 13e or 14e.

fivemack 2009-11-25 11:17

4000 @ 4e7 completed, no factor.

Use 13e. I'll take 2M-6M.

Andi47 2009-11-25 11:53

[QUOTE=fivemack;196968]4000 @ 4e7 completed, no factor.

Use 13e. I'll take 2M-6M.[/QUOTE]

reserving 6M - 7M

fivemack 2009-11-25 13:35

taking 7M-9M

Mini-Geek 2009-11-29 19:51

Line 2472 is done. Line 2473, after only some very early prelim ECM (two Quick ECMs and partial aliqueit at B1=5e4) looks like 2^5*c169.
If this c169 is semiprime, what might happen to the guide?
Edit: Never mind, found a p23, c147 remaining. Survived 2 Quick ECMs and 100@5e4 so far.
Edit 2: Finished all 214@5e4, (25 digits) will also finish the 430@25e4 (30 digits) and do some P+-1.
Edit 3: Finished 430@25e4. (30 digits) Also did P-1, max B1 100M, and P+1, max B1 50M. All in DB. Not working on it any more ATM.

jrk 2009-11-29 20:50

[code]Using B1=43000000, B2=388112953420, polynomial Dickson(30), sigma=1315477237
Step 1 took 138222ms
Step 2 took 77110ms
********** Factor found in step 2: 4123290685142646964541600067761
Found probable prime factor of 31 digits: 4123290685142646964541600067761
Composite cofactor 83203877652763456216344080057649507874810835149655593949890433813292280209307613162194128906010127548847166175620653 has 116 digits
[/code]

I will do the c116

Andi47 2009-11-29 21:04

[QUOTE=jrk;197327][code]Using B1=43000000, B2=388112953420, polynomial Dickson(30), sigma=1315477237
Step 1 took 138222ms
Step 2 took 77110ms
********** Factor found in step 2: 4123290685142646964541600067761
Found probable prime factor of 31 digits: 4123290685142646964541600067761
Composite cofactor 83203877652763456216344080057649507874810835149655593949890433813292280209307613162194128906010127548847166175620653 has 116 digits
[/code]

I will do the c116[/QUOTE]

43e6 - opening eggs with nuclear warheads? :wink:

I have found the same c31 with B1=3e6 after 68 curves ~5 minutes ago. (I planned to run it overnight, but I stopped ECMing now, hence you started GNFS)

jrk 2009-11-29 21:54

c116 = p37.p37.p43

p37 = 3134160372909851723612253077515467493
p37 = 3636840079737787436938569640509488477
p43 = 7299584610785326987792600494916936890870973

jrk 2009-11-29 23:25

I'm doing the c109 now.

jrk 2009-11-30 03:41

[QUOTE=jrk;197332]I'm doing the c109 now.[/QUOTE]

Done.

[code]
prp54 factor: 481989758002837826944000482973711890329883965855795003
prp56 factor: 16617620933533935114464391289780146092392771482229686557

[/code]

jrk 2009-11-30 06:01

line 2476 c121 finished:

[code]
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=3159156024
Step 1 took 8056ms
Step 2 took 4090ms
********** Factor found in step 2: 13432838311461612250464777552146733002407
Found probable prime factor of 41 digits: 13432838311461612250464777552146733002407
Probable prime cofactor 130982517230425503437968877198264838192630336209630023456287838747078505224065181 has 81 digits

[/code]

jrk 2009-11-30 06:12

Now, a c164.

[code]2477. 229836343198830452904906926072687726445274504040590457858093954259846740197541450378930115385305580707404985478037205666706877074351759707192765877028347973091075810594208 = 2^5 * 13 * 34031 * c164
[/code]

Batalov 2009-11-30 07:16

We have a moment of truth.
2^5 * c169

henryzz 2009-11-30 07:25

[quote=Batalov;197349]We have a moment of truth.
2^5 * c169[/quote]
that is by far larger than this forum has done
can we do it?

Batalov 2009-11-30 07:30

Let's hope it cracks, but if not... there's nothing to it :whistle: --
Womack+mersenneforum had done a gnfs-180, as you know.

henryzz 2009-11-30 07:32

[quote=Batalov;197354]Let's hope it cracks, but if not... there's nothing to it :whistle: --
Womack+mersenneforum had done a gnfs-180, as you know.[/quote]
i meant the aliquot forum
we tend to attract smaller hitters

jrk 2009-11-30 07:33

[QUOTE=henryzz;197353]that is by far larger than this forum has done
can we do it?[/QUOTE]

First, a lot of ECM.

10metreh 2009-11-30 18:13

Anyone like the idea of doing it in the Factoring forum to get more attention? (First I would have to get mod powers there, but I'll only ask for them if I need to.)

Meanwhile I'll run some curves on it at 3e6 (guessing SB has already done t35)...

BTW, I think we would need to ECM to about t55. OK?

henryzz 2009-11-30 20:36

[quote=10metreh;197381]Anyone like the idea of doing it in the Factoring forum to get more attention? (First I would have to get mod powers there, but I'll only ask for them if I need to.)

Meanwhile I'll run some curves on it at 3e6 (guessing SB has already done t35)...

BTW, I think we would need to ECM to about t55. OK?[/quote]
i think most the major hitters read this forum

fivemack 2009-11-30 20:37

[QUOTE=10metreh;197381]Anyone like the idea of doing it in the Factoring forum to get more attention? [/QUOTE]

I do not think that would be necessary.

Doing 2000@4e7 overnight.

10metreh 2009-11-30 20:53

OK, agreed. I had forgotten that Mersenneforum+Womack has come to an end (or so we expect).


All times are UTC. The time now is 05:24.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.