mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Aliquot Sequences (https://www.mersenneforum.org/forumdisplay.php?f=90)
-   -   Odds and ends....and class records (https://www.mersenneforum.org/showthread.php?t=11673)

Andi47 2009-08-04 10:23

[QUOTE=10metreh;183974]Out of interest, what is the largest GNFS you've done?[/QUOTE]

I guess, with "hard Cxyz" he means a composite which won't split by ECM. (But I'm curious about his largest GNFS too.)

gd_barnes 2009-08-04 12:31

[quote=10metreh;183974]Out of interest, what is the largest GNFS you've done?[/quote]

Correct, "hard" means it won't split by ECM alone and would be inefficient to run msieve (or yafu) on. Namely the amount of ECM curves that either Aliquiet.exe -or- Syd's DB runs by default...whichever happens to be greater for the length of cofactor in question. In other words, I let Aliquiet.exe run until it encounters GNFS and then I'll let Syd's DB hack away at it and if it finds a factor, I'll stop and restart Aliquiet.exe on the next index after putting the correct factorization in the ELF file.

Don't ask about GNFS. lol I haven't done any GNFS because I've had nothing but problems with it and as an admin for two prime seach projects, just don't have the personal time to dedicate to figuring out such issues (or even to continue posing questions for them) after already spending 2-3 hours at it after posing a couple of questions in this forum, which were answered reasonably. It's just that I kept getting different errors everytime I'd correct something.

Alas, I got a PM from Karsten a little while ago. He's going to send me a Windows care package of what he uses. It should include all necessary folders, everything modified as needed or instructions on exactly what to change, etc. I have the CYG or CWG thingie on my machine so that's not the issue. It gets through all of the relations but just won't find factors.

Believe it or not, I set the msieve to limit to 105 digits and I've done no GNFS on 19278 yet! That way it almost never goes to GNFS. The longest msieve has run on a very slow 1.6 Ghz machine so far has been 11 hours (one core) on one of the C100's that I encountered for 19278. It looks like the C102 that it is currently running on will run 16-20 hours. Crazy as it sounds, that's what I'm doing at the moment, which is why this is the first time I've tried to crack a "hard" C>100.

I will say one thing: If you don't know much about factoring to begin with, installing the GNFS stuff is extremely user unfriendly.



Gary

Greebley 2009-08-04 13:41

Oddly my experience was the opposite. With aliqueit, I didn't have to learn anything about ggnfs. A few questions on which lines to change in one file and it was up and running. Aliqueit has been good enough to solve c124's for me (I haven't tried higher due mostly to time issues and unhopeful sequences).

Without aliqueit though, I would agree with you. It is a slow learning process.

kar_bon 2009-08-09 22:20

just an eye-catcher
 
seq 514362:

[code]
797. 54774007141558011978068252997076650753477483251993617134897929348854871916939908 = 2^2 * 7^2 * 19 * 47 * [color=red]809[/color] * 4492799284639 * 182624065052414174365741 * 471458466985844972666218524873323471
798. 65244331251233353677332675714371474327771637340806111972762959309446601195124092 = 2^2 * 7 * 19 * [color=red]809^2[/color] * 4820597909 * 38871670116409488185353775390712595475476773499321140165887639
[/code]

Andi47 2009-08-28 18:54

2-digit cubed:

[code] 1029 . 231495364150475836544553534586087621550385708329062091066961390970945743804067744125728860552 = 2^3 * 23^3 * 283 * 44449 * 78059 * 165095641614636849667 * 14671069732418718140864010666279846895553901374490181157[/code]

from sequence 522348

kar_bon 2009-08-28 22:47

[QUOTE=Andi47;187829]2-digit cubed:

from sequence 522348[/QUOTE]

see post #56 for the highest cubes so far!

here's another sort of remarkables from seq 276:i1234

[code]
2406164941137993885146722410688929757159749303856305199137999910618479232219773684619394509425175460058349351330170
= 2 * 3 * 5 *
[b]566[/b]1545200582541702864324356927 *
[b]566[/b]5048649574629338619029305493 *
2500722589298704536579935613492765852985685104885849
[/code]

both 566... are P31!!!

Mini-Geek 2009-08-31 11:33

[code]Mon Aug 31 02:17:41 2009 factoring 195676107863988180517848073492543644940604662461972464243353598981074052753143825201209897621800399 (99 digits)
...
Mon Aug 31 02:17:43 2009 commencing square root phase
Mon Aug 31 02:17:43 2009 reading relations for dependency 1
...
Mon Aug 31 02:20:56 2009 reading relations for dependency 2
...
Mon Aug 31 02:24:09 2009 reading relations for dependency 3
...
Mon Aug 31 02:27:22 2009 reading relations for dependency 4
...
Mon Aug 31 02:30:33 2009 reading relations for dependency 5
...
Mon Aug 31 02:33:52 2009 reading relations for dependency 6
...
Mon Aug 31 02:37:12 2009 reading relations for dependency 7
...
Mon Aug 31 02:40:33 2009 reading relations for dependency 8
...
Mon Aug 31 02:43:51 2009 sqrtTime: 1568
Mon Aug 31 02:43:51 2009 prp47 factor: 50327499727490646814499045113166645927535479917
Mon Aug 31 02:43:51 2009 prp52 factor: 3888055415498875305644642541143657281099569913707947
Mon Aug 31 02:43:51 2009 elapsed time 00:26:10
[/code]AFAIK, this is the most dependencies I've had to go through for any of my GNFS runs before it found the factors. What are the odds that the factors wouldn't be found until the 8th (or, for that matter, nth) dependency? I think I remember hearing the odds for finding factors in any one dependency is about 1/2, so the probability of finding none in 7 would be 1/(2^7)=1/128, followed by another 1/2 chance for the 8th.

Andi47 2009-08-31 19:30

not record-breaking, but still a high power of 2 in sequence 522348:

[code]1106 . 117957006795443564706559806559733988151107700632356683434170620357205905182866206355340843956931920 = 2^4 * 5 * 394369 * 417577 * 971053321582270613 * 40497986982346005651289 * 227676340876500352213694894339949654644013389
1107 . 156294386185629296480055232759108742516550650821606278719469995837353299917159741475789081533732080 = 2^4 * 5 * 1325766445328643518013758521 * 1473622925217472662488070745113229239676692930636164143369513453757531
1108 . 207090061695958817836073183679912947924879527561409477894358383064358254081122019249844385990180864 = [B][COLOR="Red"]2^17[/COLOR][/B] * 19 * 37 * 53 * 16517386358098740629 * 2567301113077624954647850482736715304027540430973473053897368120281417
1109 . 249118838756610415034737610086481080945358879807690516464159461029754565636957302654912492413743936 = 2^6 * 19 * 37 * 43 * 761 * 15860963 * 10668137569456528577295052240525204527346449218621619384134621051924562139483381867
1110 . 298455576296218633160874067628449578723699678675330779642547194748889494094935979535436254046421184 = 2^6 * 19 * 41 * 167 * 7913701 * 82989433442147 * 2188740466464341 * 129486543576567074639562631 * 192585884214910798185892031291[/code]

schickel 2009-09-08 07:44

Check this out:[code]2423. 277629778585911849643287493387641122012726696503327039613225502287356249389590278653069430257049775938414576932228882069336576 = 2^9 * 3 * 11^2 * 31 * 48186764533956196688398283615279175902524378308723062671694753324001837251314628365260484401564057907199499739692485141
2424. 561569324866958059505740611625001360434833545199911511944933402353970595356216700002799529383577955875429485158372056930391808 = 2^8 * 3 * 11 * 31 * c121[/code]Turns out the 2^9 driver is pretty darn persistent.....

10metreh 2009-09-08 16:38

[QUOTE=schickel;188990]Check this out:[code]2423. 277629778585911849643287493387641122012726696503327039613225502287356249389590278653069430257049775938414576932228882069336576 = 2^9 * 3 * 11^2 * 31 * 48186764533956196688398283615279175902524378308723062671694753324001837251314628365260484401564057907199499739692485141
2424. 561569324866958059505740611625001360434833545199911511944933402353970595356216700002799529383577955875429485158372056930391808 = 2^8 * 3 * 11 * 31 * c121[/code]Turns out the 2^9 driver is pretty darn persistent.....[/QUOTE]

Clifford will be pleased :razz:

schickel 2009-09-13 03:29

Two in a row!
 
This one took two tries (479632):[code] 1248 . 117817016878785319006492302352690532758653456115423793341334544008108127079781481780624820223619471936102581760 = 2^9 * 3 * 5 * 11^2 * 31 * 1427 * 12044273 * 2984329241 * 106022145943107893 * 752058890281945109169558629348973124879433668213147687907006832509
1249 . 309836948446180204109058953115321603892913695932097324027900347710818395846990550767815026104030474468551784960 = 2^9 * 3 * 5 * 11^2 * 31 * 10755359280795267238078837071063049988507012572067494641286641191196875133540450003881431135864827520597
1250 . 814026164554055263522891249548435850227745982948873878582392731712968330907093356108168568582407337467481455616 = 2^10 * 3 * 11 * 31 * 6781 * 1747891 * 6352924861 * 8734847831977 * 1181477950582045778002740994994095486701945453792815702851118984443758859[/code]'course, if I had managed this one first, I could have broken the record twice....


All times are UTC. The time now is 02:00.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.