ECM results and reservations

swellman · 2014-05-13, 18:47

I'll take C211_137_135 for ECM to t50.

XYYXF · 2014-05-13, 20:53

ECM tests help us to find some relatively small factors. Usually we run GMP-ECM:
https://gforge.inria.fr/projects/ecm/

The most popular ECM tasks are:
t55: 18000 curves at B1 = 110M
t60: 42000 curves at B1 = 260M
t65: 70000 curves at B1 = 850M
It's strongly recommended to perform them sequentially (i.e. t60 only after t55). The numbers 55, 60, etc. estimate the size of the factors, e.g., t55 finds the most of factors up to 55-digit decimal length.

Currently yoyo@home is performing t55: http://www.rechenkraft.net/yoyo/down...m/xy/wu_status
Certain composites survived deeper ECM tests:
http://www.primefan.ru/xyyxf/status.html#work
http://www.mersenneforum.org/showthread.php?t=19352
http://www.mersenneforum.org/showthread.php?t=19353

There's also a list of composites which need ECM to become feasible GNFS targets:
http://www.mersenneforum.org/showthread.php?t=20318

Please take into account that information when choosing a number to be tested with ECM. If you find a factor fitting the Top-20 table, please let us know the parameters of the lucky factorization (B1 and sigma values).

pinhodecarlos · 2014-05-13, 21:21

Is there an ecmserver?

XYYXF · 2014-05-13, 21:34

It worked for nine years from March 2002 until May 2011. Then yoyo@home adopted some of its functions.

swellman · 2014-05-16, 19:53

ryanp found a p58. It's in the top 10, so I've requested the parameters he used.

He's running the remaining cofactor as a GNFS job.

swellman · 2014-05-20, 12:56

Someone found a p54 and fully factored this composite.

swellman · 2014-05-20, 13:58

I will take it up to t55.

XYYXF · 2014-05-20, 16:50

Maybe there's Ryan again? Unfortunately he doesn't inform us about his efforts :(

pinhodecarlos · 2014-05-20, 17:13

Quote:

Originally Posted by XYYXF

Maybe there's Ryan again? Unfortunately he doesn't inform us about his efforts :(

What about sending him an email?
rpropper at stanford.edu (Ryan Propper)
or
rpropper at cs.stanford.edu

swellman · 2014-05-20, 17:41

I occasionally trade PMs with Ryan, but I did not have his email address. Thanks. I'll contact him directly.

It appears he (or someone) is running the list of composites to t55. Which is very cool, as long as we document it.

A (dumb) question - is there an easy way to check the entire list of xyyxf composites against fdb to see if any others have been further factored over the years? I did a few cut/paste operations to find the factors discussed above, but that's not practical for almost 1200 composites.

swellman · 2014-05-21, 00:30

I heard from Ryan - thanks Carlos for his e-mail address.

He sends the following:

Quote:

I'm currently running ECM on the following others:

61360633774629616091.... (C191): 101^111 + 111^101 cofactor
28417942597163917291.... (C171): 107^111 + 111^107 cofactor
77215796253421280876.... (C200): 110^111 + 111^110 cofactor
20234429095044680925.... (C196): 89^112 + 112^89 cofactor
30346656057946828466.... (C210): 99^112 + 112^99 cofactor
23944615812601089637.... (C162): 78^113 + 113^78 cofactor
22006290398574795135.... (C185): 82^113 + 113^82 cofactor
23220853809120351143.... (C180): 85^113 + 113^85 cofactor
55073945672664882561.... (C168): 88^113 + 113^88 cofactor
45667518592748630156.... (C199): 92^113 + 113^92 cofactor
15126724016168709434.... (C205): 96^113 + 113^96 cofactor
94385034308959971307.... (C221): 100^113 + 113^100 cofactor
45981503153719989794.... (C187): 101^113 + 113^101 cofactor
12242314770708133797.... (C212): 102^113 + 113^102 cofactor
83028078013811481023.... (C198): 105^113 + 113^105 cofactor
86605203332969680264.... (C223): 107^113 + 113^107 cofactor
45455422131622754690.... (C223): 112^113 + 113^112 cofactor
97439042663037918249.... (C188): 89^114 + 114^89 cofactor
18512336433520902317.... (C174): 113^114 + 114^113 cofactor
58718611590521870571.... (C192): 79^115 + 115^79 cofactor
29886613171873547726.... (C210): 83^115 + 115^83 cofactor
35864182880741286844.... (C197): 87^115 + 115^87 cofactor
14522834247856712130.... (C300): 133^141 + 141^133 cofactor

This list does not address the other two composites mentioned earlier in this thread, so I've asked Ryan if they are his work or someone else. I assume he is running these to t55, but have also asked for confirmation.

eta: Ryan confirms via log file that he factored C222_115_98 with SNFS, and it found two factors. Just coincidence they look like ECM hits.

C212_113_103 is not Ryan's work.

2014-05-13, 18:47	#1
swellman Jun 2012 2²×13×59 Posts	ECM results and reservations I'll take C211_137_135 for ECM to t50. Last fiddled with by XYYXF on 2015-07-07 at 12:25

2014-05-13, 20:53	#2
XYYXF Jan 2005 Minsk, Belarus 620₈ Posts	ECM reservations ECM tests help us to find some relatively small factors. Usually we run GMP-ECM: https://gforge.inria.fr/projects/ecm/ The most popular ECM tasks are: t55: 18000 curves at B1 = 110M t60: 42000 curves at B1 = 260M t65: 70000 curves at B1 = 850M It's strongly recommended to perform them sequentially (i.e. t60 only after t55). The numbers 55, 60, etc. estimate the size of the factors, e.g., t55 finds the most of factors up to 55-digit decimal length. Currently yoyo@home is performing t55: http://www.rechenkraft.net/yoyo/down...m/xy/wu_status Certain composites survived deeper ECM tests: http://www.primefan.ru/xyyxf/status.html#work http://www.mersenneforum.org/showthread.php?t=19352 http://www.mersenneforum.org/showthread.php?t=19353 There's also a list of composites which need ECM to become feasible GNFS targets: http://www.mersenneforum.org/showthread.php?t=20318 Please take into account that information when choosing a number to be tested with ECM. If you find a factor fitting the Top-20 table, please let us know the parameters of the lucky factorization (B1 and sigma values). Last fiddled with by XYYXF on 2016-08-24 at 09:23

2014-05-13, 21:34	#4
XYYXF Jan 2005 Minsk, Belarus 2⁴×5² Posts	It worked for nine years from March 2002 until May 2011. Then yoyo@home adopted some of its functions. Last fiddled with by XYYXF on 2016-08-24 at 02:43

2014-05-16, 19:53	#5
swellman Jun 2012 2²·13·59 Posts	ECM Results ryanp found a p58. It's in the top 10, so I've requested the parameters he used. He's running the remaining cofactor as a GNFS job.

2014-05-20, 12:56	#6
swellman Jun 2012 2²×13×59 Posts	C212_113_103 Someone found a p54 and fully factored this composite.

2014-05-13, 21:21	#3
pinhodecarlos "Carlos Pinho" Oct 2011 Milton Keynes, UK 3·17·97 Posts	Is there an ecmserver?

2014-05-20, 13:58	#7
swellman Jun 2012 2²·13·59 Posts	C200_150_148 I will take it up to t55.

2014-05-20, 16:50	#8
XYYXF Jan 2005 Minsk, Belarus 2⁴·5² Posts	Maybe there's Ryan again? Unfortunately he doesn't inform us about his efforts :(

2014-05-20, 17:41	#10
swellman Jun 2012 2²·13·59 Posts	I occasionally trade PMs with Ryan, but I did not have his email address. Thanks. I'll contact him directly. It appears he (or someone) is running the list of composites to t55. Which is very cool, as long as we document it. A (dumb) question - is there an easy way to check the entire list of xyyxf composites against fdb to see if any others have been further factored over the years? I did a few cut/paste operations to find the factors discussed above, but that's not practical for almost 1200 composites.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Reservations	ET_	FermatSearch	226	2021-04-25 19:47
Linear algebra reservations, progress and results	wombatman	NFS@Home	2768	2018-06-02 15:06
NFS results and reservations	XYYXF	XYYXF Project	348	2017-09-08 04:38
Reservations	kar_bon	Riesel Prime Data Collecting (k*2^n-1)	129	2016-09-05 09:23
Where are my reservations	fivemack	PrimeNet	3	2016-02-08 17:58