C166 from 10^455-1? - Page 4

schickel · 2012-01-07, 23:23

Quote:

Originally Posted by chris2be8

I've got 8 cores in total, 1 6 core box and 1 2 core box. I'm getting about 2M relations in 14 hours so it may finish sieving in late January (it depends if the rate slows down later in the range and how many relations I need). Then add a few days for LA etc.

Chris K

OK, finally got there. I've got 118M+ unique relations and 126M+ unique ideals.

I've got some relations on the other sieving machine that I'm going to toss in the mix in hopes of a little better matrix. I'll give you an ETA later on today....

schickel · 2012-01-08, 05:35

False alarm!

Quote:

Originally Posted by msieve

commencing 2-way merge
reduce to 16890789 relation sets and 16635945 unique ideals
ignored 25 oversize relation sets
commencing full merge
memory use: 1909.9 MB
found 8544888 cycles, need 8524145
weight of 8524145 cycles is about 681972507 (80.00/cycle)
.....
commencing linear algebra
read 8524145 cycles
cycles contain 30396095 unique relations
read 30396095 relations
using 20 quadratic characters above 2147483270
building initial matrix
memory use: 4037.7 MB
read 8524145 cycles
matrix is 8523965 x 8524145 (2924.2 MB) with weight 907100851 (106.42/col)
sparse part has weight 664259727 (77.93/col)
filtering completed in 2 passes
matrix is 8515163 x 8511169 (2923.2 MB) with weight 906745720 (106.54/col)
sparse part has weight 664174894 (78.04/col)
matrix starts at (0, 0)
matrix is 8515163 x 8511169 (2923.2 MB) with weight 906745720 (106.54/col)
sparse part has weight 664174894 (78.04/col)
matrix needs more columns than rows; try adding 2-3% more relations

jasonp · 2012-01-08, 13:20

You are right on the edge of having enough relations. There's a small region where filtering will succeed but a little extra filtering inside the linear algebra will cause the job to fail. Add some more relations and you'll get past the danger zone. Add a few more relations after that and you can knock 10-15% off the size of the matrix, though whether that will reduce the total time to completion depends on how fast you can sieve.

Stargate38 · 2012-01-26, 22:45

Do you have an ETA? This number is #5 among the 10 smallest composites in the Repunit Factorizations:

http://homepage2.nifty.com/m_kamada/math/11111.htm

I love to see the new factors. It makes me happy.

Batalov · 2012-01-26, 23:18

Man is the artificer of his own happiness. Artifice something, dude, and be happy!

Why doesn't complete factorization of 10^397-1 make you happy instead? That is a sizeable achievement!*
__________

* John Littlewood while proof-reading a passage in a draft of a book noted once: "I wish I had said that". To his surprise, the final print said: "John Littlewood said:..." The printer's apprentice took his remark for the face value. :-) /a.f.a.i.r. from M.Gardner's book/

Stargate38 · 2012-01-27, 00:37

I like factorizations of any number, especially repunits and large numbers of the form k*2^n+x, such as 8675309*2^2154+2 which is 2*(8675309*2^2153+1) (Link for the second factor: http://www.factordb.com/index.php?id...00000486883929). Also I like finding Generalized Fermat primes, such as 1494^256+1 (Link: http://www.factordb.com/index.php?id...00000475946832).

kar_bon · 2012-01-27, 08:20

Quote:

Originally Posted by Stargate38

Also I like finding Generalized Fermat primes, such as 1494^256+1 (Link: http://www.factordb.com/index.php?id...00000475946832).

Before spending much time in finding GF primes, have a look at here.

Stargate38 · 2012-01-27, 17:52

Are there any primes of the form 626^2[sup]n[/sup]+1? I used Proth and didn't find any up to n=16. Also, Why does NewPGen crash when I try to sieve b^n+k with k=1 and b=8675310? See attatched image.

chris2be8 · 2012-01-27, 17:55

Quote:

Originally Posted by Stargate38

Do you have an ETA?

Linear algebra should finish in about 28 hours.So I should be able to post the factors on Sunday.

Chris K

PS. Is anyone working on R870? I could take that as my next challenge.

Batalov · 2012-01-27, 18:31

You may want to email M.Kamada - he doesn't read these forums, but he is in contact with half a dozen active repunit factorers. He would know. There are no reservations for these though.

http://homepage2.nifty.com/m_kamada/math/11111.htm
See 'Sources' and 'Recent Changes' sections.

Batalov · 2012-01-27, 18:57

Also, if you do want to take on a c162, then why not on a Wanted Cunningham c163? It was deliberately left for enthusiasts.

2012-01-26, 23:18	#38
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 2·47·101 Posts	Man is the artificer of his own happiness. Artifice something, dude, and be happy! Why doesn't complete factorization of 10^397-1 make you happy instead? That is a sizeable achievement!* __________ * John Littlewood while proof-reading a passage in a draft of a book noted once: "I wish I had said that". To his surprise, the final print said: "John Littlewood said:..." The printer's apprentice took his remark for the face value. :-) /a.f.a.i.r. from M.Gardner's book/ Last fiddled with by Batalov on 2012-01-26 at 23:21

2012-01-27, 00:37	#39
Stargate38 "Daniel Jackson" May 2011 14285714285714285714 2³·83 Posts	I like factorizations of any number, especially repunits and large numbers of the form k2^n+x, such as 86753092^2154+2 which is 2(86753092^2153+1) (Link for the second factor: http://www.factordb.com/index.php?id...00000486883929). Also I like finding Generalized Fermat primes, such as 1494^256+1 (Link: http://www.factordb.com/index.php?id...00000475946832). Last fiddled with by Stargate38 on 2012-01-27 at 00:54

2012-01-27, 17:52	#41
Stargate38 "Daniel Jackson" May 2011 14285714285714285714 664₁₀ Posts	Are there any primes of the form 626^2[sup]n[/sup]+1? I used Proth and didn't find any up to n=16. Also, Why does NewPGen crash when I try to sieve b^n+k with k=1 and b=8675310? See attatched image. Attached Thumbnails Last fiddled with by Stargate38 on 2012-01-27 at 17:53

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Team sieve #26: c166 from 4788:2661	jrk	Aliquot Sequences	38	2011-05-16 17:58
ECM for c166 from 4788:2661	frmky	Aliquot Sequences	36	2011-04-28 06:27
Team sieve #25: c166 from 4788:2632	jrk	Aliquot Sequences	13	2011-03-24 02:48
ECM work on 4788:2632 c166	schickel	Aliquot Sequences	8	2011-03-05 01:11
Team sieve #22: c166 from 3270:620	fivemack	Aliquot Sequences	55	2011-02-15 23:01

2012-01-08, 13:20	#36
jasonp Tribal Bullet Oct 2004 3·1,181 Posts	You are right on the edge of having enough relations. There's a small region where filtering will succeed but a little extra filtering inside the linear algebra will cause the job to fail. Add some more relations and you'll get past the danger zone. Add a few more relations after that and you can knock 10-15% off the size of the matrix, though whether that will reduce the total time to completion depends on how fast you can sieve.

2012-01-26, 22:45	#37
Stargate38 "Daniel Jackson" May 2011 14285714285714285714 1010011000₂ Posts	Do you have an ETA? This number is #5 among the 10 smallest composites in the Repunit Factorizations: http://homepage2.nifty.com/m_kamada/math/11111.htm I love to see the new factors. It makes me happy.

2012-01-27, 18:31	#43
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 2·47·101 Posts	You may want to email M.Kamada - he doesn't read these forums, but he is in contact with half a dozen active repunit factorers. He would know. There are no reservations for these though. http://homepage2.nifty.com/m_kamada/math/11111.htm See 'Sources' and 'Recent Changes' sections.

2012-01-27, 18:57	#44
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 9494₁₀ Posts	Also, if you do want to take on a c162, then why not on a Wanted Cunningham c163? It was deliberately left for enthusiasts.