Posts by Ageless

Basics

Project information:  
Create account
Your account
Teams
Download BOINC
Add-ons

Community

Participant profiles
Message boards
Questions and answers
Donations/Sponsors
Live Games

Statistics

Top countries
Top participants
Top computers
Top teams
Server Status
Other statistics

1) Message boards : : Number crunching : Very SLOW CPU
Posted 1388 days ago by Ageless
To be honest I don't have any idea why it slows down a pc if you set your prefs with "NO" for the option "Do work while computer is in use?".


That setting is set to Yes, to accommodate the CPDN model I have running as well. But with 1GB of RAM and the new preferences set to use 90% of memory at maximum, it shouldn't be a problem, memory wise.

In case of multi processor computers I guess it could be a problem with switching between cpu. In this case you shold try to bind the process to a cpu.

But it's not for sure, I don't have a multi processor machine to verify that.


Although my CPU is HT capable, I can't use it. My motherboard doesn't support it, nor does Windows 2000 SP4. So it's one CPU at all times.

2) Message boards : : Number crunching : Very SLOW CPU
Posted 1389 days ago by Ageless
Picking this one back up.

I haven't changed anything on my system, only my BOINC version (5.7.4 at the moment of writing). Chess960 is still slowing my whole system down when it runs. And it may only run for 10 minutes, but even those 10 minutes are annoying.

3) Message boards : : Number crunching : Very SLOW CPU
Posted 1421 days ago by Ageless
The Application needs 75 MB of RAM. Therefore it could be a problem for computers with less RAM. In this case you should not crunch for Chess960 when you need your computer to work with.


I've got a P4 2.8GHz with a total of 1GB or RAM. Even with CPDN SA, Einstein and Seti running, I don't take up all that memory.

Yet I concur, when a Chess result starts, my computer slows down enormously.

So Jonas, does Chess use L2 cache?

4) Message boards : : Number crunching : Unrecoverable error! Help!
Posted 1423 days ago by Ageless
<error_code>-200</error_code> means that the files are the wrong size.
This could be a problem that Jonas needs to solve, server side.

Since you managed to download 38 results without problem afterwards, I think it's solved.
5) Message boards : : Number crunching : Problems re-attaching to this project
Posted 1426 days ago by Ageless
Is it a problem of this BOINC Client 5.6.5 or does the project server a software update ?


First let me excuse myself for misspelling your name. I had all the letters right, just not in the correct order. ;)

If it's a problem with 5.6.5 I should have similar problems on other projects and I don't. Whether or not you require a server software update is a good guess. I'd say ask Rytis.

6) Message boards : : Number crunching : Problems re-attaching to this project
Posted 1426 days ago by Ageless
Jason, I had some fun time re-attaching to this project. :)

BOINC 5.6.5 (Alpha version, using it as Alpha tester)

Shows I failed to attach.
Shows my messages saying I am attached anyway.
When trying to attach again, I see I am already attached.

So I did an update on the project (it shows in the projects tab) and it immediately downloaded work. While for all the rest I am not attached. ;-)
7) Message boards : : Cafe Chess960 : 'Last person' thread missing
Posted 1451 days ago by Ageless
Rytis deleted it this afternoon.
8) Message boards : : Number crunching : Unrecoverable error
Posted 1454 days ago by Ageless
Is there any help or information on Error code 2?


Make sure you check if all the files in all the subdirectories to the BOINC directory have their read and write rights set.

Unhiding your computer(s) for us to see what it exactly said, is also helpful. The complete error message is written to the stderr.txt part of the result's ID.

9) Questions and answers : Platform-specific problems : Windows : Observation: Suspend leaves chess960 exe in tasklist
Posted 1456 days ago by Ageless
When BOINC is doing benchmarks, the science applications should stay in memory. This is done because in the past applications that don't checkpoint (write part of themselves, where they are in the crunch) had to start from zero percent again after they were killed by the benchmarking process.

Whether or not Chess checkpoints, I don't know.


Return to Chess960@Home main page

Copyright © 2010 Chess960@home