mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   NFS@Home (https://www.mersenneforum.org/forumdisplay.php?f=98)
-   -   Fast Breeding (guru management) (https://www.mersenneforum.org/showthread.php?t=20024)

swellman 2016-03-10 01:57

C225_144_86 survived 4000 curves @B1=11e7.

Dubslow 2016-03-10 12:53

Perhaps 14e and 15e/16e should have different queue management threads?

At any rate, Aliquot sequence 4788 is nearing readiness to begin GNFS on the extant C195. There are several questions for the NFS@Home gatekeepers. 1) How would this number be sieved? 15e or 16e? 2) What's everyone's best guess as to the total thread-hours it would take to run the GNFS (including poly select and post processing)? 3) If [URL="http://mersenneforum.org/showthread.php?p=428597#post428597"]I said that 0.5 t60 was more than sufficient[/URL] ECM in order to move to NFS, would you believe me? :smile:

debrouxl 2016-03-12 08:07

William: yeah, why not the couple C152.

Sean: thanks, queued :smile:

Bill: I leave such questions to the other gatekeepers, sorry :wink:

General 14e queue management note: I got an ECM hit on a near-repdigit number, up to 6 remain to be queued.

wblipp 2016-03-12 17:56

[QUOTE=debrouxl;428842]William: yeah, why not the couple C152.[/QUOTE]

These need polynomials.

C152 from [URL="http://factordb.com/index.php?id=1100000000438481008"]P171+1[/URL]

[URL="http://factordb.com/index.php?id=1100000000127304980"]P171[/URL] is the largest factor of [URL="http://factordb.com/index.php?id=1100000000438435798"]2778282614445777504541611766541^7-1[/URL]
2778282614445777504541611766541 is the largest factor of [URL="http://factordb.com/index.php?id=1100000000423432759"]310897033^13-1[/URL]
310897033 is the larger factor of [URL="http://factordb.com/index.php?query=%28293459^3-1%29%2F293458"]293459^3-1[/URL]
293459 is the larger factor of [URL="http://factordb.com/index.php?query=%287^11-1%29%2F6"]7^11-1[/URL]

----------------------------------------------

C152 from [URL="http://factordb.com/index.php?id=1100000000438763033"]P198+1[/URL]

[URL="http://factordb.com/index.php?id=1100000000128085192"]P198[/URL] is the largest factor of [URL="http://factordb.com/index.php?id=1100000000438438526"]6965709411950376883751815302407971477^7-1[/URL]
696570941195037688375181530240797 is the largest factor of [URL="http://factordb.com/index.php?id=1100000000310624238"]5231^17-1[/URL]

----------------------------------------------

C152 from [URL="http://factordb.com/index.php?id=1100000000438761028"]P155+1[/URL]

[URL="http://factordb.com/index.php?id=1100000000438410836"]P155[/URL] is the largest factor of [URL="http://factordb.com/index.php?id=1100000000438410834"]126975359699835793^13-1[/URL]
126975359699835793 is the largest factor of [URL="http://factordb.com/index.php?query=%2856782082406079903^3-1%29%2F56782082406079902"]56782082406079903^3-1[/URL]
56782082406079903 is the largest factor of [URL="http://factordb.com/index.php?query=%281435511378542612963^3-1%29%2F1435511378542612962"]1435511378542612963^3-1[/URL]
1435511378542612963 is the largest factor of [URL="http://factordb.com/index.php?query=%285490513541^3-1%29%2F5490513540"]5490513541^3-1[/URL]
5490513541 is the larger factor of [URL="http://factordb.com/index.php?query=%28128341^3-1%29%2F128340"]128341^3-1[/URL]
128341 is the larger factor of [URL="http://factordb.com/index.php?query=%2870841^3-1%29%2F70840"]70841^3-1[/URL]
70841 is the larger factor of [URL="http://factordb.com/index.php?query=%2819^7-1%29%2F18"]19^7-1[/URL]

XYYXF 2016-03-12 20:09

[b]C198_116_109[/b]
Sextic (difficulty 237): 116*(116[sup]18[/sup])[sup]6[/sup] + 11881*(109[sup]19[/sup])[sup]6[/sup] = 275541393882318627908853077852573592633 * C198

[b]C231_117_97[/b]
Sextic (difficulty 233): 117*(117[sup]16[/sup])[sup]6[/sup] + 912673*(97[sup]19[/sup])[sup]6[/sup] = 46 * C231

[b]C193_117_115[/b]
Sextic (difficulty 242): 117*(117[sup]19[/sup])[sup]6[/sup] + 1520875*(115[sup]19[/sup])[sup]6[/sup] = 1791655053501223228035370632926387798335118971568 * C193

Ryan Propper ran some ECM curves on them, possibly a few hundreds at B1 = 1G.

RichD 2016-03-13 04:30

[QUOTE=wblipp;428871]These need polynomials.

C152 from [URL="http://factordb.com/index.php?id=1100000000438481008"]P171+1[/URL]

[URL="http://factordb.com/index.php?id=1100000000127304980"]P171[/URL] is the largest factor of [URL="http://factordb.com/index.php?id=1100000000438435798"]2778282614445777504541611766541^7-1[/URL]
2778282614445777504541611766541 is the largest factor of [URL="http://factordb.com/index.php?id=1100000000423432759"]310897033^13-1[/URL]
310897033 is the larger factor of [URL="http://factordb.com/index.php?query=%28293459^3-1%29%2F293458"]293459^3-1[/URL]
293459 is the larger factor of [URL="http://factordb.com/index.php?query=%287^11-1%29%2F6"]7^11-1[/URL][/QUOTE]

A possible polynomial.
[CODE]N: 67396473742446277656256954217277625584479601302578823522164265944872775779400867146942596302578680975510512746707431516837825242311593178883042283748467
# expecting poly E from 3.93e-12 to > 4.52e-12
R0: -220977418297617293921859102531
R1: 414296186057323
A0: -2676345750520969345962727001608518548
A1: -9382263438325324390260050919830
A2: 4008662199500978758684164
A3: -13736762637666613
A4: -1679631487576
A5: 127908
SKEW: 2798764.94
# skew 2798764.94, size 8.956e-15, alpha -7.297, combined = 4.105e-12 rroots = 3[/CODE]

Dubslow 2016-03-13 10:24

[QUOTE=wblipp;427639]This one needs some test sieving - it would "normally" be too small, but it has big coefficients. Maybe it's still too small. It comes from a P189+1, but because of the structure of the P189 it is

[URL="http://factordb.com/index.php?id=1100000000825077234"]2426789^31+3162104763[/URL][/QUOTE]

[QUOTE=wblipp;427924]Since 31 = 5*6+1, the quintic and sextic are nearly the same. For

2426789^31+3162104763

They are

2426789*x^a + 3162104763 with x=2426789^b

(a,b) = (6,5) gives the sextic, (a,b)=(5,6) gives the quintic.[/QUOTE]

[QUOTE=debrouxl;428164]I have queued C189_147_41 (Sean) and C186_2426789_31_plus_3162104763 (William).
[/QUOTE]
A [URL="http://mersenneforum.org/showthread.php?p=428921#post428921"]slightly patched[/URL] yafu did some testing on a similar number, and decided on the quintic with 29 bit large primes, same as FDB. Interestingly, it also suggested sieving on the algebraic side. Test sieving seemed to indicate it would take a mere 6 days and 4 hours to sieve on my hyperthreaded Sandy Bridge quadcore.

[code]gen: ========================================================
gen: selected polynomial:
gen: ========================================================

n: 71913645299564138589141709037542312748165773089700209062155487060359689058327647089221985574771309040180365135426162272312780013624083749268713294754629736034429992166650355361920562201951182738553
# 2426789^31+2147483647, difficulty: 204.32, anorm: 1.44e+38, rnorm: 1.04e+44
# scaled difficulty: 204.32, suggest sieving algebraic side
# size = 1.918e-15, alpha = 0.401, combined = 2.137e-12, rroots = 1
type: snfs
size: 204
skew: 3.8849
c5: 2426789
c0: 2147483647
Y1: -1
Y0: 204264128952666376652802917061007970761
m: 204264128952666376652802917061007970761


***factors found***


***co-factor***
C197 = 71913645299564138589141709037542312748165773089700209062155487060359689058327647089221985574771309040180365135426162272312780013624083749268713294754629736034429992166650355361920562201951182738553

ans = 71913645299564138589141709037542312748165773089700209062155487060359689058327647089221985574771309040180365135426162272312780013624083749268713294754629736034429992166650355361920562201951182738553

bill@Gravemind⌚0515 ~/yafu/dev ∰∂ cat nfs.job
n: 71913645299564138589141709037542312748165773089700209062155487060359689058327647089221985574771309040180365135426162272312780013624083749268713294754629736034429992166650355361920562201951182738553
# 2426789^31+2147483647, difficulty: 204.32, anorm: 1.44e+38, rnorm: 1.04e+44
# scaled difficulty: 204.32, suggest sieving algebraic side
# size = 1.918e-15, alpha = 0.000, combined = 2.137e-12, rroots = 1
type: snfs
size: 204
skew: 3.8849
c5: 2426789
c0: 2147483647
Y1: -1
Y0: 204264128952666376652802917061007970761
m: 204264128952666376652802917061007970761

rlim: 18000000
alim: 18000000
lpbr: 29
lpba: 29
mfbr: 58
mfba: 58
rlambda: 2.6
alambda: 2.6
[/code]

Of course I can't check to see if this is what you're doing right now. Site seems to be down.

RichD 2016-03-13 19:37

[QUOTE=wblipp;428871]C152 from [URL="http://factordb.com/index.php?id=1100000000438763033"]P198+1[/URL]

[URL="http://factordb.com/index.php?id=1100000000128085192"]P198[/URL] is the largest factor of [URL="http://factordb.com/index.php?id=1100000000438438526"]6965709411950376883751815302407971477^7-1[/URL]
696570941195037688375181530240797 is the largest factor of [URL="http://factordb.com/index.php?id=1100000000310624238"]5231^17-1[/URL][/QUOTE]

A pretty good polynomial.
[CODE]N: 19047901410764002870000765673505438232741199243653795462800783185942334464971792378086671533041151507287594941733954612954780966742528799831112419286601
# expecting poly E from 4.25e-12 to > 4.88e-12
R0: -228253451035951968914815796741
R1: 101768370778103
A0: -173857566402536340068868338864897040
A1: 2614978593931839149386477763163
A2: -1038263010071723571120797
A3: -2899523306788675877
A4: 102133296162
A5: 30744
SKEW: 2717680.70
# skew 2717680.70, size 1.188e-14, alpha -6.707, combined = 4.870e-12 rroots = 5[/CODE]

Dubslow 2016-03-13 23:38

[QUOTE=Dubslow;428923]
Of course I can't check to see if this is what you're doing right now.[/QUOTE]

Oooof... sextic with 31 bit large primes. I've half a mind that I can finish it on my desktop faster than the former RSALS would...

debrouxl 2016-03-14 06:13

With 31-bit LPs and a sextic, the yield was quite good and the range relatively short. I must say that I didn't try other sieving parameters, considering how past sievings with large coefficients went :smile:

Dubslow 2016-03-14 08:23

[QUOTE=debrouxl;429030]With 31-bit LPs and a sextic, the yield was quite good and the range relatively short. I must say that I didn't try other sieving parameters, considering how past sievings with large coefficients went :smile:[/QUOTE]

I just changed Yafu's coefficients to use 64 bit coefficients, allowing this number as is -- and the half a bit difference in the additive coefficient changed the sextic to be better (with 29 bit lps). Still algebraic side, but now with a 5 day ETA on my box. *Increasing* the number and coefficient both, by a seemingly trivial amount, made the job substantially easier. Somewhat extraordinary to my amateur eyes...


All times are UTC. The time now is 22:58.

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