[QUOTE=akruppa;286180]Doing NFS on the c129.[/QUOTE]
I'm about 2/3 done. I should have factors overnight or in the morning...

Actually, I underestimated - it's already filtering...

Sorry, I wasn't more clear earlier when I wrote, "Working it..."

edit: Hmm, had some overlap between a couple machines and it looks like a huge percentage of duplicates.:sad: Might be back to overnight. Maybe you can get yours done before mine...

Starting P-1/P+1/ECM on the c143...

Edit: P-1 @ 1e9 done. P+1 @ 1e9 done. 1000 @ 1M done. Now ECM @ 3M

Edit 2: Factor found:

[CODE]GMP-ECM 6.4 [configured with GMP 5.0.90, --enable-asm-redc] [ECM]
Using B1=3000000, B2=5706890290, polynomial Dickson(6), 4 threads
Finished 729/3000; avg s/curve: stg1 5.309s, stg2 2.105s; runtime: 1698s

Run 729 out of 3000:
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=3028862197
Step 1 took 3764ms
Step 2 took 1564ms
********** Factor found in step 2: 2949088942941563277174422986781707688321642807
Found probable prime factor of 46 digits: 2949088942941563277174422986781707688321642807
Composite cofactor 9736626601903181478128027287780617910920260701054260773118334149516337462506185479911300273299887 has 97 digits
[/CODE]

I put 7600 ECM curves @ 43M into yoyo@home queue and send then with high priority.
yoyo

[QUOTE=yoyo;286267]I put 7600 ECM curves @ 43M into yoyo@home queue and send then with high priority.
yoyo[/QUOTE]
You can cancel them :smile:

Edit: P-1/P+1 @ 1e9 on the c97 done. 1000 @ 1M done. 250 @ 3M done.

Ok, canceled.

[QUOTE=Batalov;276134]Greg broke it, Greg fixed it! You da man![/QUOTE][QUOTE=Batalov;284521]Alex broke it, Alex fixed it! Looks like a pattern. :rolleyes:[/QUOTE]
OK, who brought in the 5s. Take them out, please.:smile:

I'll run some ecm on the C150:
[url]http://www.rechenkraft.net/yoyo/y_status_ecm.php[/url]

If you did 7000@44M and 5500@110M already, it's time to stop ECM and to start looking for a polynomial. I did another ~1000@44M here so far.

Where is [B]jrk[/B] when you need him? He comes up with some of the best polys.

[QUOTE=RichD;286325]Where is [B]jrk[/B] when you need him? He comes up with some of the best polys.[/QUOTE]

I'll post one by tomorrow

This is what I got with CADO so far, after searchin 1/4 of the proposed search length:

[CODE]
# /ufs/kruppa/cado-nfs-1.1-RC5/build/pidgey.pna.cwi.nl//polyselect/polyselect2 -q -kmax 25 -incr 30030 -admin 0 -admax 200000000 -degree 5 -maxnorm 48.7 -t 1 100000
# Initializing primes took 1ms
# Tried 6660 ad-value(s), found 889 polynomial(s), 20 below maxnorm
# potential collisions=1.13e+03 (2.20e-01/s)
# best logmu: 46.85 47.33 47.66 47.66 47.74 47.81 48.19 48.22 48.33 48.39
# Total phase took 5153.77s
n: 429277684954121394813422187749189045233252715220675732389611040151522593868963683035384720260595974399002374247125295377293926247838643737190293927843
skew: 141760.000
# lognorm: 48.07, alpha: -6.82 E=41.25
# Murphy's E(Bf=10000000,Bg=5000000,area=1.00e+16)=4.16e-12
c5: 112552440
c4: 56498047731994
c3: 10636367348344460659
c2: -2154667295744034144611234
c1: -46795479420098489335085286606
c0: -6358821065971945895700364097496303
Y1: 1763652201127
Y0: -20714606119241218423970191838
m: 392491080660317669180028037964838401073066155249179927872282843937679271571986439895186505836807460649157814882710296198845337728044091927343534129827
# generated by b325e6e: /ufs/kruppa/cado-nfs-1.1-RC5/build/pidgey.pna.cwi.nl//polyselect/polyselect2 -q -kmax 25 -incr 30030 -admin 0 -admax 200000000 -degree 5 -maxnorm 48.7 -t 1 100000 in 5153.77s
[/CODE]

Judging by the other polynomials' Murphy score, this seems to be a pretty far outlier; the second best polynomial has a score of 3.5e-12. I'll finish the other 3/4 of the search.

Edit: finished the search but did not find anything better.

current best: (using latest msieve GPU)
[code]
Edit: deleted as per request
[/code]
Although I expect closer to 6e-12 with more effort. But if anyone wants to start on sieving, be my guest.

[QUOTE=akruppa;286331]This is what I got with CADO so far, after searchin 1/4 of the proposed search length:
[CODE]# Murphy's E(Bf=10000000,Bg=5000000,area=1.00e+16)=4.16e-12
[/CODE]...
[/QUOTE]
msieve gauges size 1.160e-14, alpha -6.822, combined = 4.607e-12 rroots = 1 for this one. (Just to be sure to compare apples to apples. If cado-nfs will produce a 5.4e-12, that would be as good as 6e-12 in msieve.)

[code]
Edit: deleted as per request
[/code]

edit: still running poly search for several more hours but feel free to start sieving this poly. RSALS?

[code]# sieve with ggnfs lasieve4 I14e on alg side from Q=8M to ??? (not completed)
# aq4788:2864
n: 429277684954121394813422187749189045233252715220675732389611040151522593868963683035384720260595974399002374247125295377293926247838643737190293927843
# norm 1.792057e-14 alpha -8.037412 e 6.038e-12 rroots 1
skew: 1306475.69
c0: 214908009559587306593785542266030152
c1: 929214552266747520591256982462
c2: 253789725595123464400862
c3: 9328486332206420599
c4: -927067067574
c5: 292824
Y0: -68112450209164394866751341767
Y1: 43236795456461179
rlim: 16000000
alim: 16000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.5
alambda: 2.5[/code]

mods can please delete the earlier two polys I posted if they desire.

Well, RSALS is currently on a low activity period...

Let's say I can queue it (as usual, starting at the bottom of the range) tomorrow morning if fivemack, bsquared or others are not willing to do a unilateral sieve by then ?

Sadistic remark: Now depending of how the C150 splits, there is a nice chance to get my favorite driver, 2^3*3*5 ....
(and I can hear the echoes: Booo!!! Shut up!! Go to hell!! etc... :yucky:)

The c150 is -1 (mod 4), so no, not at this iteration.

[QUOTE=Batalov;286440]The c150 is -1 (mod 4), so no, not at this iteration.[/QUOTE]
You may be right, I believe you. (I did not check it, I only wanted to see the reaction of the guys around, throwing their pencils, mouses, whatever, at me through the monitor... :P, but don't tell them!)

[QUOTE=LaurV;286420]...favorite driver, 2^3*3*5 ....
[/QUOTE]

I'd prefer 2[SUP]1[/SUP]*3[SUP]0[/SUP]

OK, starting "C150_4788_2864" on RSALS :smile:
For now, I've queued q=8M-20M, but what would be the expected minimum upper bound for q values ?

[QUOTE=debrouxl;286459]OK, starting "C150_4788_2864" on RSALS :smile:
For now, I've queued q=8M-20M, but what would be the expected minimum upper bound for q values ?[/QUOTE]

I'll do the post-processing.

[QUOTE=debrouxl;286459]OK, starting "C150_4788_2864" on RSALS :smile:
For now, I've queued q=8M-20M, but what would be the expected minimum upper bound for q values ?[/QUOTE]

I just tested it. Use 8M-32M for the range

In the meantime I run some more ECM curves ;)
[url]http://www.rechenkraft.net/yoyo/y_status_ecm.php[/url]
But no factor yet :(
yoyo

At the time of this writing, ~42M raw relations returned for "C150_4788_2864", range 8M-33M, with an estimated remaining ~14.5M raw relations.
The dumb linear projection is always optimistic, so I guess we'll have about 55M raw relations at the end of the sieving.

The .fb and .ini files are not on this computer, but anyway, I'm just using the latest poly posted by jrk several posts above, the one with e 6.038e-12.

[QUOTE=debrouxl;286582]

The .fb and .ini files are not on this computer, but anyway, I'm just using the latest poly posted by jrk several posts above, the one with e 6.038e-12.[/QUOTE]

:tu:

By the way, I am still 20 hours away to have my computer free for C150_4788_2864 post-processing, and by looking at the RSALS page, it is almost ready for testing so I can give up my reservation if anyone is interested.

I take it, ETA is 11 hours.

[QUOTE=unconnected;286621]I take it, ETA is 11 hours.[/QUOTE]

Ok, it's yours!

prp70 factor: 1853984410157827166179951247265224756997707661816330288396414485621389
prp81 factor: 231543308887682378366790127665185278502890914525881554486862581961175850876303087

Unfortunately it picks up [B]3^2[/B].

Running gnfs on c117@i2872.

Doing curves at 11M on the c149. Edit: 44M

[QUOTE=akruppa;286711]Doing curves at 11M on the c149.[/QUOTE]
running poly search. I'll post one tonight unless someone finds a factor.

[QUOTE=jrk;286717]running poly search. I'll post one tonight unless someone finds a factor.[/QUOTE]

[code][color=blue]# sieve with ggnfs lasieve4 I14e on alg side from Q=8M to 30M[/color]
# aq4788:2875
n: 44944968120798970101956399260161048307127624608363627768855259786875744937516514253381492976598354377631809096637742242405056022967763594463388725947
# norm 2.165061e-14 alpha -8.021700 e 6.736e-12 rroots 5
skew: 2076131.15
c0: -2724152092669553301350919758264155248
c1: 19148983504629007377598273822658
c2: -6876114514422853669377235
c3: -12263406826598895700
c4: 947045644148
c5: 794352
Y0: -35524959071412568874372278765
Y1: 10439608423446181
rlim: 16000000
alim: 16000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.5
alambda: 2.5
[/code]

[QUOTE=Andi47;286445]I'd prefer 2[SUP]1[/SUP]*3[SUP]0[/SUP][/QUOTE]
2^0*x is even better, for any x except a square... :P

Found 6836077383647882167495637621893086479456936569

********** Factor found in step 2: 4387050301986034667561975512302171805093213

Update - another one:

********** Factor found in step 2: 583358317624551201719447527348636163

Number 3:

Found probable prime factor of 44 digits: 10174073890280464531539919236320115459913679

The sequence tries to imitate 890460 with it's 2^9 · 3^8 :smile::

2^2 · 3^4 after 2^2 · 3^2 and 2^2 · 3.

Now a c136 @ i2883.

[QUOTE=LaurV;286754]2^0*x is even better, for any x except a square... :P[/QUOTE]

What would it require to get 2^0*x? something like 2²*p²?

[QUOTE=Andi47;286847]What would it require to get 2^0*x? something like 2²*p²?[/QUOTE]
2[SUP]m[/SUP]z[SUP]2[/SUP], m>0.
[URL="http://www.factordb.com/sequences.php?se=1&eff=2&aq=2%5E60*%282%5E61-1%29*112986307451471003597&action=range&fr=0&to=10"]Example[/URL].

I guess one of my machines found something overnight for i2883:
[code]
[Jan 20 2012, 23:52:04] c136: running 1218 ecm curves at B1=11e6...
Curve 2:
Using B1=11000000, B2=35133391030, polynomial Dickson(12), sigma=4074077776
Step 1 took 181411ms
Step 2 took 59136ms
********** Factor found in step 2: 504473322692355276852610423345559
[/code]

We lost the 3! :fusion:

[QUOTE=akruppa;287067]We lost the 3! :fusion:[/QUOTE]
Ok, now all we need is to loose the "1" from the "17" that appears the last three terms (.2900-.2902) :sirrobin:

[QUOTE=LaurV;287084]Ok, now all we need is to loose the "1" from the "17" that appears the last three terms (.2900-.2902) :sirrobin:[/QUOTE]

Don't wake a sleeping bear.
:bear:

btw, I'd prefer losing one or two specimens of that even prime.

c147: 1000@11e6

c147: 3000@1e6; 3500@11e6

c147: +1000@11e6 (total 2000)

I will run polynomial selection overnight.

3600@44M

c147: +1000@11e6 (total 3000)

In case it matters, c147 +1000 more @ 11e6 (total 4000)

I will evaluate tomorrow, whether I can sieve at any appreciable rate with any of my current machines. I've been too slow in the past...

Attempting 7M-8M to see what happens...

edit: BTW +1000 more @ 11e6 (total 5000)

Oops. I forgot to report the p52 from iteration 2846 to Paul Zimmermann for his 2012 top ten ECM factors list. I guess it will drop out of the top ten in a few days.

OK, I completed 7M-8M, turning up ~1.65M relations. I'll take 8-10M now to play with further.

I have several machines still running ecm @ 11e6 and have totalled up ~7500. At this point is B1=11e6 a futile search?

Yes, B1=11M is futile now.

The [strike]c144[/strike] (Edit: that's [B]c147[/B]) is done. The next 3 lines fell to ECM easily. Currently at i2906 on a [URL="http://factordb.com/index.php?id=1100000000488319639"]c144[/URL]. I've completed 904 B1=1M curves. I'm not planning to run B1=3M curves right now.

I have turned loose several slow cores @11e6...

I queued some with 43e6 and 110e6.
status -> [url]http://www.rechenkraft.net/yoyo/y_status_ecm.php[/url]

1800@44M

4788:2906-c144: 1000@11e6

4788:2906-c144: +1000@11e6 (2000 total)

4788:2906-c144: +1000@11e6 (3000 total)

4788:2906-c144: +1000@11e6 (4000 total)

4788:2906-c144: +1000@11e6 (5000 total)

Is this probably enough @11e6?

With the yoyo curves, even B1=44M should be futile now. This one is over-due for NFS.

[QUOTE=akruppa;288778]With the yoyo curves, even B1=44M should be futile now. This one is over-due for NFS.[/QUOTE]
Yeah, I was thinking of pointing my machines in a different direction, but I don't know if that should be this effort or something else. I can at least supply some help here for a c144 and 4788 interest has seemed to be weak as of late. So, if a "good" poly shows up, I'll do some sieving...

I'm doing poly selection. Maybe post one tonight or in the morning.

[code][color=blue]# sieve with ggnfs lasieve4 I14e on alg side from Q=5M to 20M[/color]
# aq4788:2906
n: 216210261026078873728038575619824007502275880651339130269087415140753033343108746166779571643387473335848998664028620971224681169067812545897739
# norm 7.636505e-14 alpha -7.107237 e 1.402e-11 rroots 3
skew: 662015.51
c0: 1110267212789719902085595616638897
c1: 166651926133804056753801800942
c2: 504288034858202176634199
c3: -330129543228280610
c4: -1301252763020
c5: 601392
Y0: -3244471923903548991075971280
Y1: 4996695193542667
rlim: 10000000
alim: 10000000
lpbr: 28
lpba: 28
mfbr: 56
mfba: 56
rlambda: 2.55
alambda: 2.55
[/code]

reserving 5M-8M

It's weekend, I can steal CPU time over at CWI. I'll sieve 8M up.

My range is finished and the files are up @ SendSpace:

[URL="http://www.sendspace.com/file/tv4rz7"]4788-2906a.zip (292.53MB)[/URL]

[URL="http://www.sendspace.com/file/yhcr7q"]4788-2906b.zip (90.03MB)[/URL]

Done.

i2910:c148: 1710 @ 3e6

4000@11M. Doing 44M now.

1000 @ 11e6

running poly selection on line 2910. I will post a poly sunday night or monday morning (EST).

I've done about 5k@44M, this should be ready for NFS.

1k@1e6 on 2911, 2k@1e7 running tonight

on 2911
ran 1 p-1 with B1=1e8, B2=5638436970138,
ran 1 p+1 with B1=1e8, B2=6083968236318,
nothing found

p-1: B1=1e10, B2=1e15, no factor found

2k@44M

6400@11M on the c129 of 2915.

7600@43M on the c129 of 2915 queued, 7110 done
18000@110M on the c129 of 2915 queued, 16250 done
20000@260M on the c129 of 2915 queued, 11700 done
20000@850M on the c129 of 2915 queued, 195 done

Status here: [url]http://www.rechenkraft.net/yoyo/y_status_ecm.php[/url]

I named the 43M, 110M, 260M wrong, but these are the c129 of 2915. I think the composite is now ready for sieving?

yoyo

[QUOTE=yoyo;292167]7600@43M on the c129 of 2915 queued, 7110 done
18000@110M on the c129 of 2915 queued, 16250 done
20000@260M on the c129 of 2915 queued, 11700 done
20000@850M on the c129 of 2915 queued, 195 done[/QUOTE]
Running NFS. I'll post factors thursday afternoon or evening (cst).

[QUOTE=yoyo;292167]7600@43M on the c129 of 2915 queued, 7110 done
18000@110M on the c129 of 2915 queued, 16250 done
20000@260M on the c129 of 2915 queued, 11700 done
20000@850M on the c129 of 2915 queued, 195 done

Status here: [url]http://www.rechenkraft.net/yoyo/y_status_ecm.php[/url]

I named the 43M, 110M, 260M wrong, but these are the c129 of 2915. I think the composite is now ready for sieving?[/QUOTE]

You have done two orders of magnitude more work than would be involved in simply doing the sieving! A C129 is an 80-hour job, a single curve at 260M takes half an hour and you've done more than ten thousand of them - this ECM barrage is comparable to the sieving effort required for a C160.

[QUOTE=fivemack;292177]You have done two orders of magnitude more work than would be involved in simply doing the sieving! A C129 is an 80-hour job, a single curve at 260M takes half an hour and you've done more than ten thousand of them - this ECM barrage is comparable to the sieving effort required for a C160.[/QUOTE]

I wonder how the yoyo BOINC members would feel to know that but they are clueless of the math involved. They just run for the stats...lol

[QUOTE=yoyo;292167]7600@43M on the c129 of 2915 queued, 7110 done
18000@110M on the c129 of 2915 queued, 16250 done
20000@260M on the c129 of 2915 queued, 11700 done
20000@850M on the c129 of 2915 queued, 195 done

Status here: [url]http://www.rechenkraft.net/yoyo/y_status_ecm.php[/url]

I named the 43M, 110M, 260M wrong, but these are the c129 of 2915. I think the composite is now ready for sieving?

yoyo[/QUOTE]
It seems like the yoyo@home have found a p58 of the c129 with B1=85e7, and is it possible to show the sigma of the factor on the table please?

I hoped to find the sigma through [url]http://www.rechenkraft.net/yoyo/download/download/stats/ecm/[/url] , but there doesn't seem to be a folder for Aliquot sequences...

[QUOTE=fivemack;292177]You have done two orders of magnitude more work than would be involved in simply doing the sieving! A C129 is an 80-hour job, a single curve at 260M takes half an hour and you've done more than ten thousand of them - this ECM barrage is comparable to the sieving effort required for a C160.[/QUOTE]

Where can I see which amount of curves on which B1 level should be done before sieving?

yoyo

[QUOTE=yoyo;292190]Where can I see which amount of curves on which B1 level should be done before sieving?[/QUOTE]

For GNFS, sieve to 1/3 of the composite size. For a C129 that would be 43 digits. To see how many curves that takes, run ecm with -v option. With B1=11e6, 40 digits is 681 curves and 45 digits is 4480 curves. Interpolating gives 2960 curves, so about 3000 at 11e6 is enough. Two or three times this might be justified on the basis of convenience.

[QUOTE=debrouxl;292189]I hoped to find the sigma through [url]http://www.rechenkraft.net/yoyo/download/download/stats/ecm/[/url] , but there doesn't seem to be a folder for Aliquot sequences...[/QUOTE]

Now there is a as folder with the result and the sigma:
[url]http://www.rechenkraft.net/yoyo/download/download/stats/ecm/as/FOUND/C129_4788_i2915/ecm_as_1331056199_C129_4788_i2915_17540[/url]

yoyo

[QUOTE=wreck;292181]It seems like the yoyo@home have found a p58 of the c129 with B1=85e7, and is it possible to show the sigma of the factor on the table please?[/QUOTE]
The era of NFS misses has come! There are now entirely too many p65s for c172-s, p58 for c176-s... then why not a p58 for the c129.

2k@11M on the c138.

2400@11e6 on the c138, no factor

The c138 has been factored as p66 * p73 (not by me... I wonder if it was also done with ecm :smile:).

The current status is i2920 with a c136.

[QUOTE=rajula;292505]The c138 has been factored as p66 * p73 (not by me... I wonder if it was also done with ecm :smile:).
[/QUOTE]

Nope, I did it by NFS. A first try of such size on my I7.

Now we have a c153 on line 2924.

So far I have done 600 ECM curves @ B1=8e7, B2=9.7e11

And there is this polynomial:
[code][color=blue]# sieve with ggnfs lasieve4 I14e on alg side from Q=5M to 33M[/color]
# est ~55M raw relations (avg. 0.061 sec/rel C2D @ 3.4GHz)
# aq4788:2924
n: 112296381884896012490614382451451986588162409728169933283326536718943569566088868056285056843610277229580864777319128827749822341401205107820881104153761
# norm 1.007367e-14 alpha -5.944168 e 4.262e-12 rroots 5
skew: 581194.32
c0: 105458850334534013736584980745623680
c1: 261196250419483240838309407992
c2: -1631204900348006739703582
c3: -2536445228874907803
c4: 6724024379652
c5: 856596
Y0: -167308382383766648362675345607
Y1: 59894083073807711
rlim: 20000000
alim: 20000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.5
alambda: 2.5
[/code]
I tested six polynomials with murphy score from 4.209e-12 to 4.472e-12 and the above polynomial had the best yield.

I am already sieving with this poly:

[CODE]# norm 1.179679e-14 alpha -7.474522 e 4.832e-12 rroots 5
skew: 4546142.80
c0: -6064723996191997325826878478448793241
c1: 36018107245891279042997152353345
c2: -8649683750492276196117189
c3: -5758528824171653297
c4: 305864195158
c5: 64680
Y0: -280490396538732147442124580146
Y1: 788153256539269661
[/CODE]

I expect to have the factors on sunday.

Congrats!!! May your down driver last longer than the last ones did! :grin::popcorn::toot::fusion:

:tu: