Advanced search

Message boards : Graphics cards (GPUs) : 6.55 windows application - error by error!

Author Message
Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 5045 - Posted: 29 Dec 2008 | 16:29:16 UTC
Last modified: 29 Dec 2008 | 16:46:44 UTC

A few last tasks for 6.55 application - all of them has "Computation error":

29.12.2008 22:00:39|GPUGRID|Sending scheduler request: To fetch work. Requesting 1 seconds of work, reporting 0 completed tasks
29.12.2008 22:00:44|GPUGRID|Scheduler request completed: got 1 new tasks
29.12.2008 22:00:46|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-LICENSE
29.12.2008 22:00:46|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-COPYRIGHT
29.12.2008 22:00:47|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-LICENSE
29.12.2008 22:00:47|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-COPYRIGHT
29.12.2008 22:00:47|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-eb17700-SH2_USPME-1-40-SH2_USPME930000_1
29.12.2008 22:00:47|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-eb17700-SH2_USPME-1-40-SH2_USPME930000_2

29.12.2008 22:01:24|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-eb17700-SH2_USPME-1-40-SH2_USPME930000_2
29.12.2008 22:01:24|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-eb17700-SH2_USPME-1-40-SH2_USPME930000_3
29.12.2008 22:01:26|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-eb17700-SH2_USPME-1-40-SH2_USPME930000_3
29.12.2008 22:01:26|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-complex_full.sol.ionized.pdb

29.12.2008 22:01:31|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-eb17700-SH2_USPME-1-40-SH2_USPME930000_1
29.12.2008 22:01:31|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-complex_full.sol.ionized.psf
29.12.2008 22:03:09|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-complex_full.sol.ionized.pdb
29.12.2008 22:03:09|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-parameters
29.12.2008 22:03:16|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-parameters
29.12.2008 22:03:16|GPUGRID|Started download of eb17700-SH2_USPME-2-40-SH2_USPME930000-SH2_USPME930000
29.12.2008 22:03:43|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-complex_full.sol.ionized.psf
29.12.2008 22:04:02|GPUGRID|Finished download of eb17700-SH2_USPME-2-40-SH2_USPME930000-SH2_USPME930000
29.12.2008 22:04:03|GPUGRID|Starting eb17700-SH2_USPME-2-40-SH2_USPME930000_4
29.12.2008 22:04:03|GPUGRID|Starting task eb17700-SH2_USPME-2-40-SH2_USPME930000_4 using acemd version 655
29.12.2008 22:04:06|GPUGRID|Computation for task eb17700-SH2_USPME-2-40-SH2_USPME930000_4 finished
29.12.2008 22:04:06|GPUGRID|Output file eb17700-SH2_USPME-2-40-SH2_USPME930000_4_1 for task eb17700-SH2_USPME-2-40-SH2_USPME930000_4 absent
29.12.2008 22:04:06|GPUGRID|Output file eb17700-SH2_USPME-2-40-SH2_USPME930000_4_2 for task eb17700-SH2_USPME-2-40-SH2_USPME930000_4 absent
29.12.2008 22:04:06|GPUGRID|Output file eb17700-SH2_USPME-2-40-SH2_USPME930000_4_3 for task eb17700-SH2_USPME-2-40-SH2_USPME930000_4 absent
29.12.2008 22:04:08|GPUGRID|Started upload of eb17700-SH2_USPME-2-40-SH2_USPME930000_4_0
29.12.2008 22:04:11|GPUGRID|Finished upload of eb17700-SH2_USPME-2-40-SH2_USPME930000_4_0

http://www.gpugrid.net/result.php?resultid=188071

What's wrong? I'll try to restart my machine, but... There are already several tasks of such kind...

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 123
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 5046 - Posted: 29 Dec 2008 | 17:06:52 UTC - in response to Message 5045.

Which drivers are you using?
180.84 should fix the "out of memory" error on XP64.
____________

pixelicious.at - my little photoblog

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 5053 - Posted: 29 Dec 2008 | 19:23:01 UTC

I've updated my drivers. Now

30.12.2008 1:20:22|GPUGRID|Sending scheduler request: To fetch work. Requesting 505048 seconds of work, reporting 0 completed tasks
30.12.2008 1:20:27|GPUGRID|Scheduler request completed: got 0 new tasks
30.12.2008 1:20:27|GPUGRID|Message from server: No work sent
30.12.2008 1:20:27|GPUGRID|Message from server: Full-atom molecular dynamics for Cell processor is not available for your type of computer.
30.12.2008 1:20:27|GPUGRID|Message from server: Full-atom molecular dynamics on Cell processor is not available for your type of computer.


BOINC Manager v6.5.0

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 5062 - Posted: 30 Dec 2008 | 2:28:32 UTC

Got one. Seems to be OK. Thanx for an advice!

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 5385 - Posted: 8 Jan 2009 | 16:37:09 UTC

Another 6 faulty WUs!

http://www.gpugrid.net/result.php?resultid=204325

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 5390 - Posted: 8 Jan 2009 | 18:47:31 UTC

Seems that BOINC 6.5.0 can't make a gpugrid WU run normally while there's one or more lhc@home's WUs being processed...

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 123
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 5394 - Posted: 8 Jan 2009 | 19:23:03 UTC - in response to Message 5390.

Don't think so... I also use BOINC 6.5.0 and I'm running GPUGRID and LHC right now. No problems so far...
____________

pixelicious.at - my little photoblog

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5395 - Posted: 8 Jan 2009 | 19:56:29 UTC - in response to Message 5390.

Seems that BOINC 6.5.0 can't make a gpugrid WU run normally while there's one or more lhc@home's WUs being processed...


Me too ...

Two systems ...

Sadly, I can't seem to grab that much work off LHC ... even though I have been trying for ages ... sigh ... my favorite project ...

I guess I best not think about it or I will be getting even more depressed ...

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5397 - Posted: 8 Jan 2009 | 22:01:55 UTC - in response to Message 5390.

Seems that BOINC 6.5.0 can't make a gpugrid WU run normally while there's one or more lhc@home's WUs being processed...


I have made the same experience in the past, so I only crunch LHC on my other boxes without CUDA. Don't know why there are problems, but know that ... :(

____________

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 5411 - Posted: 9 Jan 2009 | 5:12:24 UTC - in response to Message 5394.

Don't think so... I also use BOINC 6.5.0 and I'm running GPUGRID and LHC right now. No problems so far...


Well... You are lucky guy, what else can we say here... :-) Hope that BOINC crew will deal with it...

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5426 - Posted: 9 Jan 2009 | 19:14:33 UTC - in response to Message 5394.

Don't think so... I also use BOINC 6.5.0 and I'm running GPUGRID and LHC right now. No problems so far...


@Stefan: Do you run Linux? I have no problems with LHC and BOINC on Linux, but many problems on Windows 64 bit.

____________

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 123
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 5428 - Posted: 9 Jan 2009 | 21:13:31 UTC - in response to Message 5426.
Last modified: 9 Jan 2009 | 21:14:19 UTC

Don't think so... I also use BOINC 6.5.0 and I'm running GPUGRID and LHC right now. No problems so far...


@Stefan: Do you run Linux? I have no problems with LHC and BOINC on Linux, but many problems on Windows 64 bit.


Actually I use both, Linux and Windows.

The Vista 64 bit host with BOINC 6.5.0 x86_64 is crunching a mix of LHC, GPUGRID and a few other projects without any problems. The same with Linux 64 bit - no problems if I run LHC + GPUGRID. But on the Linux hosts I still use BOINC 6.4.2. Because they are only remote controlled with BOINCview and without monitor and keyboard/mouse I don't update them very often...
____________

pixelicious.at - my little photoblog

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5441 - Posted: 10 Jan 2009 | 12:46:32 UTC - in response to Message 5411.

Hope that BOINC crew will deal with it...


I don't think it's a BOINC problem, it's got something to do with the science apps and the graphics driver.

A question to all you guys who have problems with GPu-Gird and LHC: are you watching the LHC graphics? Even once may be enough to trigger the error, if their app doesn't free the vid memory correctly.

MrS
____________
Scanning for our furry friends since Jan 2002

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5448 - Posted: 10 Jan 2009 | 14:17:02 UTC

In that I love my LHC I don't plan to try ...

However, I will say that I have two machines happily running both LHC and GPU Grid ... LHC work is validating and running to completion normally ... but I do not use the graphics except on rare occasions ...

The screen saver mode kicking on may also be an issue for those that use that ... or allow that ...

Fun times ...

I posted this on the BOINC Dev mailing list:


There are scattered reports that running GPU Grid at the same time as LHC@Home can cause failure of LHC tasks.

For those experiencing this problem we are asking for more input to see if we can discover more about the problem.

Current suspicion is that LHC may not be correctly clear the video memory. Thus if the participant looks at the graphics or the screen saver kicks on there will be an issue.

This may be an issue actually with the BOINC Graphics library interface where it does not do proper garbage collection or state restoration.

Post to thread

Message boards : Graphics cards (GPUs) : 6.55 windows application - error by error!

//