mersenneforum.org  

Go Back   mersenneforum.org > Math Stuff > Computer Science & Computational Number Theory

Reply
 
Thread Tools
Old 2020-02-02, 23:21   #23
Stargate38
 
Stargate38's Avatar
 
"Daniel Jackson"
May 2011
14285714285714285714

2·13·23 Posts
Default

@Paulunderwood: Any chance you could post a FactorDB link for that number, or at least post a zipped .txt attachment with the full decimal expansion? I tried clicking "Show all of the digits", but I just got this:

Code:
Partition numbers are too slow to calculate 'on the fly' and should be pre-calculated. Tell the editor this one is
missing and it will be added

Last fiddled with by Stargate38 on 2020-02-02 at 23:22 Reason: formatting
Stargate38 is offline   Reply With Quote
Old 2020-02-02, 23:45   #24
paulunderwood
 
paulunderwood's Avatar
 
Sep 2002
Database er0rr

CBF16 Posts
Default

Quote:
Originally Posted by Stargate38 View Post
@Paulunderwood: Any chance you could post a FactorDB link for that number, or at least post a zipped .txt attachment with the full decimal expansion? I tried clicking "Show all of the digits", but I just got this:

Code:
Partition numbers are too slow to calculate 'on the fly' and should be pre-calculated. Tell the editor this one is
missing and it will be added
The top5000 uses an outdated version of Par/GP to calculate the digits of partition numbers. My version takes a minute or two -- the version the top5000 is using has already taken over 12 hours.

I am just uploading the 79MB certificate to FactorDB. Verification will tie up a core for several days. Somehow factorDB rejected the certificate saying "candidate must be probable prime". It will only be a matter of a couple of days before the certificate appears on ellipsa.eu.

I have almost completed a verification of the certificate here on my 48 core -- it is likely to be around 20 hours in total.

See the attached for the decimal expansion.
Attached Files
File Type: zip numbpart_1289844341.txt.zip (19.6 KB, 29 views)

Last fiddled with by paulunderwood on 2020-02-03 at 00:33
paulunderwood is offline   Reply With Quote
Old 2020-02-03, 01:33   #25
Batalov
 
Batalov's Avatar
 
"Serge"
Mar 2008
Phi(3,3^1118781+1)/3

5×1,811 Posts
Default

Congrats!

Here is a method that I use to create large entries in FactorDB.
1. Find any unrelated composite.
2. Paste your number in the 'Report factors' text box. ;-)
3. Get the link to the number on the next webpage. It is http://factordb.com/index.php?id=1100000001443762221

Another method is to submit the zipped primo's out file and then find it in your own
http://factordb.com/certoverview.php?userid=${your-uid} but when it is large then what you described likely happens.
Batalov is offline   Reply With Quote
Old 2020-02-03, 02:23   #26
paulunderwood
 
paulunderwood's Avatar
 
Sep 2002
Database er0rr

1100101111112 Posts
Default

Quote:
Originally Posted by Batalov View Post
Congrats!

Here is a method that I use to create large entries in FactorDB.
1. Find any unrelated composite.
2. Paste your number in the 'Report factors' text box. ;-)
3. Get the link to the number on the next webpage. It is http://factordb.com/index.php?id=1100000001443762221

Another method is to submit the zipped primo's out file and then find it in your own
http://factordb.com/certoverview.php?userid=${your-uid} but when it is large then what you described likely happens.
Thanks for creating the number. I have tried 3 different ways to upload the cert. and am giving up.

I have verified the number in about 15.5 hours on my 48 core.

I await Marcel putting up the cert. to which I will link from UTM.
paulunderwood is offline   Reply With Quote
Old 2020-02-03, 07:51   #27
Batalov
 
Batalov's Avatar
 
"Serge"
Mar 2008
Phi(3,3^1118781+1)/3

5·1,811 Posts
Default

When I did a bit of an exploration of Lehmer numbers, I've done those depositions a lot.
I also typically initialize N-1 and N+1, in this case
http://factordb.com/index.php?id=1100000001443795933
http://factordb.com/index.php?id=1100000001443795971

The interesting thing is that for them, the PRP test didn't choke so FDB can properly treat the entries of this size.
Batalov is offline   Reply With Quote
Old 2020-02-03, 10:07   #28
paulunderwood
 
paulunderwood's Avatar
 
Sep 2002
Database er0rr

1100101111112 Posts
Default

The cert has been uploaded to FDB -- it had to change from status "U" to "PRP" before I could upload successfully.
paulunderwood is offline   Reply With Quote
Old 2020-02-03, 10:11   #29
axn
 
axn's Avatar
 
Jun 2003

11×421 Posts
Default

Quote:
Originally Posted by paulunderwood View Post
The cert has been uploaded to FDB -- it had to change from status "U" to "PRP" before I could upload successfully.
Not seeing it, though. Neither under the number nor in the queue (http://factordb.com/certoverview.php...p=0&pending=on)
axn is online now   Reply With Quote
Old 2020-02-03, 10:26   #30
paulunderwood
 
paulunderwood's Avatar
 
Sep 2002
Database er0rr

13×251 Posts
Default

Quote:
Originally Posted by axn View Post
Not seeing it, though. Neither under the number nor in the queue (http://factordb.com/certoverview.php...p=0&pending=on)
I don't know why it is not showing up in that queue. The certificate is now on Ellipsa top20.

I am still awaiting UTM to register it so I can link to there from the ECPP Wiki page.

Last fiddled with by paulunderwood on 2020-02-03 at 10:29
paulunderwood is offline   Reply With Quote
Old 2020-02-03, 20:36   #31
ewmayer
2ω=0
 
ewmayer's Avatar
 
Sep 2002
Rep├║blica de California

52·11·41 Posts
Default

Nice work on these, Paul - catching up with the thread:

Quote:
Originally Posted by paulunderwood View Post
It is with great pleasure that I proudly announce a new record ECPP certification of a 34093 digits prime. Arranged in a 331 by 103 grid the digits of the prime depicts the Mersenne prime exponent 57,885,161.

The certification process took 14 months using a 4 x AMD 6174 (48 cores at 2.2 GHz) and 200 days additionally spent in parallel on an AMD 1090T (6 cores at 3.2 GHz) running phase 2 work.
I see the trailing digits = ...000532669 - how many such candidates did you have to PRP-test to find this one? Your above snip mentions just the ECPP effort, if I read that correctly.
ewmayer is offline   Reply With Quote
Old 2020-02-03, 21:36   #32
paulunderwood
 
paulunderwood's Avatar
 
Sep 2002
Database er0rr

13·251 Posts
Default

Quote:
Originally Posted by ewmayer View Post
Nice work on these, Paul - catching up with the thread:



I see the trailing digits = ...000532669 - how many such candidates did you have to PRP-test to find this one? Your above snip mentions just the ECPP effort, if I read that correctly.
That little touch of a prime ending would have been done like this: start with the picture ending zeroes and have pari/gp add primes to the number to form a list in a file -- maybe with some elementary trial division. Then process the file with pfgw64 -f until a PRP shows. As I recollect I had one or two found overnight.

Further, the big digits of the picture were created in text files and stitched together with Perl.

Last fiddled with by paulunderwood on 2020-02-03 at 21:42
paulunderwood is offline   Reply With Quote
Old 2020-02-11, 05:26   #33
paulunderwood
 
paulunderwood's Avatar
 
Sep 2002
Database er0rr

326310 Posts
Default

Quote:
Originally Posted by paulunderwood View Post
I am still awaiting UTM to register it so I can link to there from the ECPP Wiki page.
After nearly 9 days UTM has given "External" status to the partitions number. I can now do the Wiki page.
paulunderwood is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
ECPP-DJ danaj Computer Science & Computational Number Theory 58 2018-02-06 21:52
Can I just leave this here? (ECPP) trhabib Miscellaneous Math 6 2011-08-19 16:34
Fast ECPP T.Rex Math 6 2007-06-25 16:42
Where ECPP is needed? nuggetprime Information & Answers 9 2007-04-06 16:22
new ECPP article R. Gerbicz GMP-ECM 2 2006-09-13 16:24

All times are UTC. The time now is 04:01.

Sat Jul 4 04:01:09 UTC 2020 up 101 days, 1:34, 1 user, load averages: 1.44, 1.32, 1.22

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.