mersenneforum.org  

Go Back   mersenneforum.org > Other Stuff > Archived Projects > 15k Search

 
 
Thread Tools
Old 2005-10-20, 15:19   #56
TTn
 

100001011100012 Posts
Default

That's interesting, I had already made it so it is selected, but I didn't think of having the full form ready to input, for manual use.
This is easy todo.


Allright quit upstaging me. lol
Shane F.
 
Old 2005-10-20, 15:53   #57
Flatlander
I quite division it
 
Flatlander's Avatar
 
"Chris"
Feb 2005
England

31×67 Posts
Default

25) Issue. At the moment clicking the icon moves/shrinks the form.
Flatlander is offline  
Old 2005-10-20, 15:57   #58
TTn
 

2×3×5×7×47 Posts
Default

Oh no, the world is coming to an end, and the sky is falling!

I noticed the resize inconsistency.
 
Old 2005-10-20, 17:58   #59
Flatlander
I quite division it
 
Flatlander's Avatar
 
"Chris"
Feb 2005
England

31×67 Posts
Default

I asked for that.

When approaching the end of the candidates, is it worth forcing newpgen to run at 1% CPU (or even close) even if more sieving is 'requested'. Seems silly to sieve for hours, then llr, when it would be quicker to just llr.

ie. (pseudo-code)

If (estimatedtimeleft) < (timer*1.5) then close(newpgen);
Flatlander is offline  
Old 2005-10-21, 04:32   #60
TTn
 

3×7×193 Posts
Default opt

Quote:
When approaching the end of the candidates, is it worth forcing newpgen to run at 1% CPU (or even close) even if more sieving is 'requested'. Seems silly to sieve for hours, then llr, when it would be quicker to just llr.
ie. (pseudo-code)
If (estimatedtimeleft) < (timer*1.5) then close(newpgen);
I understand your question, but don't know how you're applying it.
Approching the end, LLR would be at 99%, forcing NewPGen to 1%.
Only if sieving is faster at eliminating candidates, will it be requested at 99%.
It would be silly to test, when it's quicker to sieve, and vice versa. Right?

The psedudo code, gives me a better idea of your question.
Do you mean only at the end, or at the "point of no return", where NewPgen is completely not expected to come back?

The LLR finish time could be estimated, to force the last timer loop's with NewPGen closed. So that should be ok, but will only save <%1.
The point of no return, is hard with fixed k, since the variation is unknown.
This could be an optimization if acurately done, early enough for %1 to add up over time. This must not make a mistake though, and close Newpgen to early.
 
Old 2005-10-21, 11:01   #61
TTn
 

4,993 Posts
Default

I've been running a large RMA.NET test for about 65 hours now.
k= 2145
n = 250000 thru 1250000
One million candidates of fixed k, RMAing.

NewPGen rate 1192 (dominant process)
LLR rate 200
Time period: 8 1/2 hours
 
Old 2005-10-21, 19:27   #62
Flatlander
I quite division it
 
Flatlander's Avatar
 
"Chris"
Feb 2005
England

31·67 Posts
Default

Running fixed k = 31395 for n=205000 to 315000 for about 5 hours. NewPGen is dominant as expected, timer at 256.

Does 'Verbose' work for you?
Flatlander is offline  
Old 2005-10-21, 21:06   #63
TTn
 

22·5·191 Posts
Default verbose bug

You got me dead to rights.
This is my bug, I have two menu Verifies.
One of which writes Verbose=1 in the llr.ini file.
It should be hooked to it's proper menu option for Verbose.
I missed this proof reading the code, because verify and verbose looked the same.

Nice find!
I think Jean intended to now use this, for more of a diagnostic purpose.
Would'nt help much, if it does'nt work correctly due to me... hehe
 
Old 2005-10-21, 21:11   #64
TTn
 

26×53 Posts
Default

I've Stopped a large RMA.NET test at about 68.19444 hours now.
k= 2145
n = 250000 thru 1250000
One million candidates of fixed k, RMAing.

NewPGen rate 723 (dominant process)
LLR rate 200
Time period: 8 1/2 hours


I'll resume this later, after the bulk of 0.9 is completed.
 
Old 2005-10-23, 16:58   #65
Flatlander
I quite division it
 
Flatlander's Avatar
 
"Chris"
Feb 2005
England

31·67 Posts
Default

I am delighted to announce the birth of RMA.NET 0.8's firstborn.

After 2 days of labour came the birth of a slightly underweight '31395*2^208615-1'

Gifts of money may be sent via PayPal to my email account

Last fiddled with by Flatlander on 2005-10-23 at 16:59
Flatlander is offline  
Old 2006-01-02, 02:50   #66
TTn
 

52·17 Posts
Default RMA.NET Version 0.83T

RMA.NET version 0.83T is now available.
We need help testing the software on different target machines, before a release version is made.

The file can be found here, or downloaded as a group zip.
http://15k.us/aboutus.html
 
 

Thread Tools


All times are UTC. The time now is 05:50.

Tue Aug 4 05:50:57 UTC 2020 up 18 days, 1:37, 0 users, load averages: 1.57, 1.35, 1.28

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.