mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > PrimeNet

Reply
 
Thread Tools
Old 2017-03-03, 01:06   #1211
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

32·13·61 Posts
Default

Quote:
Originally Posted by ramgeis View Post
As I TF them only to 70 (as intended) they are waiting to expire (should happen soon).
I looked into this. To avoid lots of "churn", the server assigns minimum bit lengths based on exponent size so that each test takes at least a few minutes to complete. Exp < 50M, minimum 70 bits, < 100M 71 bits, < 200M 72 bits, < 400M 73 bits, >400M 74 bits.

I'm willing to listen to arguments for adjusting this policy.
Prime95 is offline   Reply With Quote
Old 2017-03-03, 12:22   #1212
ramgeis
 
ramgeis's Avatar
 
Apr 2013

22·29 Posts
Default

Quote:
Originally Posted by Prime95 View Post
I looked into this. To avoid lots of "churn", the server assigns minimum bit lengths based on exponent size so that each test takes at least a few minutes to complete. Exp < 50M, minimum 70 bits, < 100M 71 bits, < 200M 72 bits, < 400M 73 bits, >400M 74 bits.

I'm willing to listen to arguments for adjusting this policy.
Interesting, my current proceeding is quite the opposite. When I request a TF assignment from a range close to the one where the LL-action is going on I manually change the factoring limit to the highest number needed in order to have the exponent ready for the next stage. My impression is that having the TFing finished for an exponent in that range as soon as possible is more useful than working on another one and not finishing it either. On the other hand in the higher ranges the focus is more in having a good factor-finding throughput to eliminate as many exponents as soon as possible and this is more likely while working on lower bit levels on many exponents.
ramgeis is offline   Reply With Quote
Old 2017-03-03, 16:58   #1213
Mark Rose
 
Mark Rose's Avatar
 
"/X\(‘-‘)/X\"
Jan 2013

23×359 Posts
Default

Quote:
Originally Posted by ramgeis View Post
Interesting, my current proceeding is quite the opposite. When I request a TF assignment from a range close to the one where the LL-action is going on I manually change the factoring limit to the highest number needed in order to have the exponent ready for the next stage. My impression is that having the TFing finished for an exponent in that range as soon as possible is more useful than working on another one and not finishing it either. On the other hand in the higher ranges the focus is more in having a good factor-finding throughput to eliminate as many exponents as soon as possible and this is more likely while working on lower bit levels on many exponents.
That's basically what GPU72 does...
Mark Rose is online now   Reply With Quote
Old 2017-03-03, 22:21   #1214
chalsall
If I May
 
chalsall's Avatar
 
"Chris Halsall"
Sep 2002
Barbados

221068 Posts
Default

Quote:
Originally Posted by ramgeis View Post
My impression is that having the TFing finished for an exponent in that range as soon as possible is more useful than working on another one and not finishing it either.
This is true. It is most useful to get candidates TF'ed to (currently) 76 bit before they are handed off to the P-1'ers. But the truth is the TF'ing is currently several years ahead of the LL'ers, but only a few months ahead of the P-1'ers.

Quote:
Originally Posted by ramgeis View Post
On the other hand in the higher ranges the focus is more in having a good factor-finding throughput to eliminate as many exponents as soon as possible and this is more likely while working on lower bit levels on many exponents.
Please remember that this is the Great Internet Mersenne PRIME Search.

Finding factors may be amusing and satisfying, but unless they are found just before the LL'ing wavefronts they don't really add much to the main project.

Having said that, many have invested a lot of time, effort and money finding factors on not yet tested candidates. Much easier (read: less expensive) gratification.
chalsall is offline   Reply With Quote
Old 2017-03-04, 05:23   #1215
LaurV
Romulan Interpreter
 
LaurV's Avatar
 
Jun 2011
Thailand

212278 Posts
Default

Quote:
Originally Posted by Prime95 View Post
I'm willing to listen to arguments for adjusting this policy.
Don't!
LaurV is offline   Reply With Quote
Old 2017-04-10, 15:32   #1216
nordi
 
Dec 2016

3×13 Posts
Default

A few days ago, I use mprime to run a P-1 factorization of M1283 with this line in my worktodo:

Pminus1=N/A,1,2,1283,-1,7974098357,637927868521,40,"4824675346114250541198242904214396192319"

The server responded with
Quote:
[Comm thread Apr 4 09:25] Sending result to server: UID: nordi/nordiputer, M1283 completed P-1, B1=7974098357, B2=637927868521, E=12, We8: 62D7DE0A
[Comm thread Apr 4 09:25]
[Comm thread Apr 4 09:25] PrimeNet error 40: No assignment
[Comm thread Apr 4 09:25] P-1 result for M1283 was not needed
[Comm thread Apr 4 09:25] Done communicating with server.

The "No assignment" part is correct. I hand-edited my worktodo, I was checking if I had the syntax right for numbers with known factors.

But why is the result "not needed"? According to mersenne.ca/exponent/1283, both my B1 and B2 are larger than those of previous P-1 runs.
nordi is offline   Reply With Quote
Old 2017-04-10, 15:50   #1217
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

32·13·61 Posts
Default

It is a bit confusing.

The PrimeNet server does not keep track of P-1 results for partially factored Mersenne numbers. Thus, the result not needed message.

James' web site does keep track of such P-1. The PrimeNet server recorded your result, James will download it (daily?), and presumably update his web site.
Prime95 is offline   Reply With Quote
Old 2017-04-10, 16:02   #1218
nordi
 
Dec 2016

3×13 Posts
Default

Quote:
Originally Posted by Prime95 View Post
James' web site
Is that mersenne.ca?
I sent the result 6 days ago, yet it is still not shown. Other results I sent were synced to mersenne.ca in roughly a day, as you wrote.

I also should have mentioned that this result is not shown on my results page on mersenne.org.
nordi is offline   Reply With Quote
Old 2017-04-10, 16:10   #1219
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

3·1,039 Posts
Default

Quote:
Originally Posted by Prime95 View Post
James' web site does keep track of such P-1. The PrimeNet server recorded your result, James will download it (daily?), and presumably update his web site.
The data is automatically sync'd between mersenne.org -> mersenne.ca every overnight.

However, I'm concerned that perhaps PrimeNet is not, in fact, recording that result. This result is several days old and does not appear on (my view of) the mersenne.org exponent history. Some time ago the manual results page was updated to permit handling of this kind of situation, so I think it would be handled as described if the result was submitted manually. Perhaps the PrimeNet API has a different codepath for such results and also needs to be rewritten? Can you look into that please George?
James Heinrich is online now   Reply With Quote
Old 2017-04-10, 16:25   #1220
GP2
 
GP2's Avatar
 
Sep 2003

1010000101012 Posts
Default

Quote:
Originally Posted by nordi View Post
A few days ago, I use mprime to run a P-1 factorization of M1283 with this line in my worktodo:

Pminus1=N/A,1,2,1283,-1,7974098357,637927868521,40,"4824675346114250541198242904214396192319"

...

But why is the result "not needed"? According to mersenne.ca/exponent/1283, both my B1 and B2 are larger than those of previous P-1 runs.
I have already done this exponent M1283 using P−1 with B1=100000000000, B2=106381511725640448 without finding any new factor. I have also done many other very small exponents, however stage 2 for these was done with GMP-ECM rather than mprime, so they can't be reported automatically to PrimeNet, and I can't be 100% sure that there isn't some error in the script I wrote to convert from mprime savefile format to GMP-ECM savefile format that might have caused it to miss factors in rare cases.
GP2 is offline   Reply With Quote
Old 2017-04-10, 17:12   #1221
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

3·1,039 Posts
Default

Quote:
Originally Posted by GP2 View Post
I have already done this exponent M1283 using P−1 with B1=100000000000, B2=106381511725640448 without finding any new factor. I have also done many other very small exponents, however stage 2 for these was done with GMP-ECM rather than mprime, so they can't be reported automatically to PrimeNet
If you have them available, could you please email me a complete results log for all said P-1 on "very small" exponents and I'll see about manually stuffing those results into at least the mersenne.ca database.
James Heinrich is online now   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Official "Faits erronés dans de belles-lettres" thread ewmayer Lounge 39 2015-05-19 01:08
Official "all-Greek-to-me Fiction Literature and Cinema" Thread ewmayer Science & Technology 41 2014-04-16 11:54
Official "Lasciate ogne speranza" whinge-thread cheesehead Soap Box 56 2013-06-29 01:42
Official "Ernst is a deceiving bully and George is a meanie" thread cheesehead Soap Box 61 2013-06-11 04:30
Official "String copy Statement Considered Harmful" thread Dubslow Programming 19 2012-05-31 17:49

All times are UTC. The time now is 00:53.

Sat Oct 24 00:53:56 UTC 2020 up 43 days, 22:04, 0 users, load averages: 0.92, 1.03, 1.17

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.