mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   NFS@Home (https://www.mersenneforum.org/forumdisplay.php?f=98)
-   -   BOINC NFS sieving - NFS@Home (https://www.mersenneforum.org/showthread.php?t=12388)

wombatman 2014-12-01 14:19

GC_7_288 splits into 3 factors as:

[CODE]prp65 factor: 27228980019063723477666908210565377804689617876318193719310944931
prp67 factor: 7995558297218444662787725295273354129158887910319249395421003564723
prp74 factor: 26867294369158921791926370813716017769039617417668710594893300002355040827
elapsed time 184:21:20[/CODE]

Matrix density was 125, matrix size was 11.5M, and 4 threads were used.

VictordeHolland 2014-12-01 14:48

GC_6_307 factors
 
1 Attachment(s)
[B]GC_6_307[/B]
[code]
prp90 factor: 258481544812146546801940027474293767410382801030601321177479034651779093939406957125376707
prp147 factor: 256648336418042842273662200039142464906438220889736003376447339726160318944778280965564378768668904347862528786256410432396088108668220816221253557
[/code]
13M matrix with target_density 110

VictordeHolland 2014-12-01 15:07

reserving GC_9_257b
 
taking [B]GC_9_257b[/B]

swellman 2014-12-03 16:25

Reserving GC_3_514
 
If it is still available.

VictordeHolland 2014-12-06 22:02

[QUOTE=Batalov;388527]Cofactor of [URL="http://factordb.com/index.php?query=I%281345%29%2FI%28269%29%2F5"]F1345[/URL] could be a nice target.
The polynomial is a quartic which should be propped with uneven fb lims - the patterns for large quartics are easily found in older quartic projects. Then, maybe, F1405 and F1415 - using the same poly. All three have no known factors and are well-ECMd.

[COLOR=DarkGreen]EDIT: P.S. Because there was no interest, I'll run F1345 myself.[/COLOR][/QUOTE]
It's not that I'm not interested, the thing is that Greg is the only one with access to the management pages for lasievee (v5) and he is rather busy at the moment.

I also emailed the factorizations of L1277, L1279 and L1908 to the Lucas/Fibonacci pages (mersennus.net) and asked him to credit them with NFS@home. They were completed a few months ago by NFS@home, but not yet listed on the page.

Yul, please also mail Marin (of mersennus.net) once the factorization of L1291 is complete, he is not following this thead.

VictordeHolland 2015-01-14 21:16

[QUOTE=Mini-Geek;392387]
Would I be correct in assuming that today's lasieved 31 bit (~210M relations) numbers take significantly longer to postprocess (e.g. 7 days vs 1 days) than the numbers that were common here a few months ago (mostly 30 bit, IIRC)? I had to lower my target density (from 120 to 110), and was surprised at how long the ETA for this number was, until I realized that most of my earlier postprocs had much fewer relations.[/QUOTE]
A 12.7M matrix with target_density=110 is comparable to my latest 14e NFS post-processing. There are occasional outliers with more relations/smaller matrices:

Number SNFS(diff) U-relations(M) [I]Target_Density[/I] [B]Matrix (M)[/B]
GC_2_795 SNFS(245) 183 [I]120[/I] [B]10.6[/B]
GC_3_503 SNFS(245) 191 [I]120[/I] [B] 9.4[/B]
GC_6_309 SNFS(245) 181 [I]110[/I] [B]9.6[/B]
GC_3_506 SNFS(245) 200 [I]128[/I] [B]9.5[/B]
GC_3_509 SNFS(245) 175 [I]110[/I] [B]13.2[/B]
GC_8_272 SNFS(245) 193 [I]120[/I] [B]9.7[/B]
GC_6_306 SNFS(245) 228 [I]130[/I] [B]7.1[/B]
GC_6_307 SNFS(245) 176 [I]110[/I] [B]13.0[/B]
GC_9_257 SNFS(245) 170 [I]110[/I] [B]13.4[/B]
GC_8_274 SNFS(245) 172 [I]110[/I] [B]11.9[/B]
GC_6_313 SNFS(245) 164 [I]105[/I] [B]15.0[/B]

fivemack 2015-01-16 11:37

OK, I've switched to aiming for 230M relations for the SNFS(245ish) numbers, in the hope this will make the matrices a bit smaller. GW_12_226 (with 220M) is a 14M matrix which will take me around six days on 4 cores i7/4770.

The GNFS numbers are using low-balled relation targets (to the point that I'd not expect target densities much over 70 to work) but even a C170 GNFS is a good deal easier than an SNFS245; I'm trying to get a better idea of where the 31/32 crossover is.

pinhodecarlos 2015-01-27 11:04

Good job guys. In the meantime I already been asked for a new NFS@Home challenge (SETI.USA team). Is it possible to add a stats page inside NFS@Home page for challenges? We can easily make one 15e challenge...just a thought.

Carlos

pinhodecarlos 2015-01-27 16:42

They want a BOINC challenge, that's it. They don't care which application (siever).

xilman 2015-02-19 21:05

Just added GC_4_409 and GC_8_264 with difficulty 249 and 241 respectively to the 14e queue.

All the other SNFS targets under S250 are either reserved to NFS@Home, reserved to Rob Hooft for pre-testing or under a t50 on ECM.

The following are presently all 15 unreserved GNFS possibilities < C165 with at least a t50 done, some of them well over a t55. Unfortunately no polynomials are yet available but perhaps a gpu-day each would suffice. The gpu-msieve here is presently broken following a CUDA upgrade but I'll try to mend it again. Even if successful, the polynomials won't be ready in time for the Gadarene rush so anyone else who would like to join in is most welcome!

[code]
726963666889756270439831124137652409588166392285606791962251219514352986840188163853738237343537308637132102683314584642847874085671459624004763984549261 GW_6_345
279206957436231362416243740347350360954719777858809484217761790782778696037661871148414798786123246612331739045504573805572494305905017406469870061147908029 GW_6_349
365481454562518298386703403690854358624085574953056305919891375148887703537673009500976640261594452288197251807999104096705213227313896461393396768161048729 C_981
444220122038226339148094949293574156667054866866219710644112533751192418587542607220912745680701834676577594103479217938041694745147156644444249331483389359 GC_3_635
723084371423926628450019264871360448248146235409687853942826013366527897308889676983110470190478971003947335924871598925281385272136096307085354720249264851 GW_8_337
2549270044379441417530956143103979269602530372250240774544131665947778230233355051606633194688361285997719902440101095379072448186746312845362283910566112067 W_925
11564735537909688699361377694250362295685697201920782941561272354978319056467313183922643887916848573717103978651337208909702781734689978170193898421918189479 GC_8_251
15245685683654194070528451784367735927134564102584982532265254072253746820313011591722538299058918175786099377638450142276578178842074686064451121444334808941 W_862
44735593203455368013457292321310587598015229402228910880250269370420404714500756480164816920559868056101726025956894722487436154893385257223678874474018051363 GC_4_409
7406947131808134323216215757753034874585607301853060335348284949930069413805840372534786972924476987602597404701289051389140243745408221962473492776680481330727 GW_6_289
12157961952891457571864160991125659725320057642276069457092927876094586327572247029776767146204228968234216228642288911251567673823002334235997239751844277149623 GW_3_598
25435347625490803254414761928223085633430542136719026623079992763795469819811367056562821693686831136596215092561436265263385489921420902123463528883017929030511 GC_9_274
75803210067186787906592420218497285841721679059828266745898521112776907671837510357219312394505119703064118263621901096904357009101084801500737467351521680744051 W_847
429029423121896674206591410060656286271312034858546242336901864226665117427625764458140583950338797912894203455691819016559421597936003179046137489332768988628331 GW_3_615
3033875390746706032246711855799676210454874561023806340362922037043514338095101078059920839874529400044563420081700179851170691903873792537159584161556560209667519 GW_4_576
[/code]

RichD 2015-02-25 02:57

GC_3_635 poly
 
Here is a nice one:

[CODE]# expecting poly E from 2.31e-12 to > 2.65e-12
R0: -1359198285270137128371654381319
R1: 59019921238571
A0: 1417602298009801357429076392205672742600
A1: 1539532967769496186334886927331994
A2: 142741045485670568846592131
A3: -43468850102584704076
A4: -714911776044
A5: 95760
skew 10489821.20, size 5.193e-15, alpha -9.158, combined = 3.020e-12 rroots = 5[/CODE]

I'll work GW_8_337 next.


All times are UTC. The time now is 21:53.

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