Deprecated: Function get_magic_quotes_gpc() is deprecated in /srv/BOINC/live-webcode/html/inc/util.inc on line 640
BRP v 1.31 feedback

WARNING: This website is obsolete! Please follow this link to get to the new Albert@Home website!

BRP v 1.31 feedback

Message boards : Problems and Bug Reports : BRP v 1.31 feedback
Message board moderation

To post messages, you must log in.

AuthorMessage
Alex

Send message
Joined: 1 Mar 05
Posts: 88
Credit: 398,734
RAC: 0
Message 112265 - Posted: 6 Nov 2012, 22:13:34 UTC

got 4 ati and one cuda.
cuda finished without any problem.
I tried to run 2 wu's on both ati gpu's. Well, one of them is the on-chip apu. This caused the system to crash; bluescreen pointed to an amd-driver problem.
Switching back to one wu per gpu solved the problem. Anyway, not a problem of the application.
3 of five are finished, waiting for validation.
ID: 112265 · Report as offensive     Reply Quote
Profile Viking69
Avatar

Send message
Joined: 2 Mar 05
Posts: 5
Credit: 2,024,604
RAC: 0
Message 112266 - Posted: 8 Nov 2012, 4:13:49 UTC

I now have 3 WU's on my only PC that hase a CUDA GPU. One is a 1.28 and 2 are 1.31.

see them here

I see NO progress with these. If I suspend all other work on this PC I do not see that these unit start to process, either as a CPU app or a GPU app. I monitor my cores and GPU temperature, and after I suspended all other owrk all my cores and the GPU go to cool temperatures. This tells me that nothing is going on.

If I look at my profile I only see these WU's and it looks like mone of my other PC's are participating.

Win 7 x64, 16GB Ram, Nvidia GTX550 Ti ver 310.33
ID: 112266 · Report as offensive     Reply Quote
Profile Viking69
Avatar

Send message
Joined: 2 Mar 05
Posts: 5
Credit: 2,024,604
RAC: 0
Message 112267 - Posted: 8 Nov 2012, 16:02:51 UTC - in response to Message 112266.  
Last modified: 8 Nov 2012, 16:02:59 UTC

I now have 3 WU's on my only PC that hase a CUDA GPU. One is a 1.28 and 2 are 1.31.

see them here

I see NO progress with these. If I suspend all other work on this PC I do not see that these unit start to process, either as a CPU app or a GPU app. I monitor my cores and GPU temperature, and after I suspended all other owrk all my cores and the GPU go to cool temperatures. This tells me that nothing is going on.

If I look at my profile I only see these WU's and it looks like mone of my other PC's are participating.

Win 7 x64, 16GB Ram, Nvidia GTX550 Ti ver 310.33


I noticed that my othe GPU WU's were also having this problem. I restarted BOINC fully and the WU's started processing.
ID: 112267 · Report as offensive     Reply Quote
Profile Stephan Goll

Send message
Joined: 13 Dec 05
Posts: 19
Credit: 1,874,367
RAC: 0
Message 112271 - Posted: 9 Nov 2012, 19:32:35 UTC

The scheduler seems to go wild.

381513 145690 9 Nov 2012 | 18:00:24 UTC 9 Nov 2012 | 19:28:30 UTC Error while computing 1.37 0.05 --- Binary Radio Pulsar Search v1.00 (BRP3cuda32fullCPU)
381169 145518 9 Nov 2012 | 16:24:28 UTC 9 Nov 2012 | 19:28:30 UTC Error while computing 1.37 0.05 --- Binary Radio Pulsar Search v1.00 (BRP3cuda32fullCPU)
380960 145159 9 Nov 2012 | 13:11:17 UTC 9 Nov 2012 | 16:24:28 UTC Error while computing 2.87 0.07 --- Binary Radio Pulsar Search v1.00 (BRP3cuda32fullCPU)
380430 145199 9 Nov 2012 | 10:49:22 UTC 9 Nov 2012 | 13:11:17 UTC Completed and validated 5,796.33 288.62 500.00 Binary Radio Pulsar Search v1.31 (BRP3cuda32nv270)

I got one cuda270 workunit and it was fine. Then I got cuda32full workunits and they all failed.
Stephan
ID: 112271 · Report as offensive     Reply Quote
Profile Bikeman (Heinz-Bernd Eggenstein)
Volunteer moderator
Project administrator
Project developer
Avatar

Send message
Joined: 28 Aug 06
Posts: 1483
Credit: 1,864,017
RAC: 0
Message 112272 - Posted: 11 Nov 2012, 16:46:30 UTC - in response to Message 112271.  

The scheduler seems to go wild.

381513 145690 9 Nov 2012 | 18:00:24 UTC 9 Nov 2012 | 19:28:30 UTC Error while computing 1.37 0.05 --- Binary Radio Pulsar Search v1.00 (BRP3cuda32fullCPU)
381169 145518 9 Nov 2012 | 16:24:28 UTC 9 Nov 2012 | 19:28:30 UTC Error while computing 1.37 0.05 --- Binary Radio Pulsar Search v1.00 (BRP3cuda32fullCPU)
380960 145159 9 Nov 2012 | 13:11:17 UTC 9 Nov 2012 | 16:24:28 UTC Error while computing 2.87 0.07 --- Binary Radio Pulsar Search v1.00 (BRP3cuda32fullCPU)
380430 145199 9 Nov 2012 | 10:49:22 UTC 9 Nov 2012 | 13:11:17 UTC Completed and validated 5,796.33 288.62 500.00 Binary Radio Pulsar Search v1.31 (BRP3cuda32nv270)

I got one cuda270 workunit and it was fine. Then I got cuda32full workunits and they all failed.
Stephan


Oh, sorry... that 1.00 BRP3cuda32fullCPU version was supposed to be deprecated (deactivated) on the server side. Ignore those. If you have any of those left, feel free to abort them.

Thanks for reporting this.

Cheers
HBE
ID: 112272 · Report as offensive     Reply Quote
Profile Holmis

Send message
Joined: 4 Jan 05
Posts: 104
Credit: 2,104,736
RAC: 0
Message 112273 - Posted: 11 Nov 2012, 22:00:31 UTC

Hi

So far the BRP v1.31 has ran quite nicely on my host (id 2267).
But I do have two WU that looks a bit interesting from a testing stand point, namely:

WU id: 143172 with 3 inconclusive so far, 2 CUDA and one OpenCL

and

WU id: 147263 with 2 inconclusive, both CUDA
ID: 112273 · Report as offensive     Reply Quote

Message boards : Problems and Bug Reports : BRP v 1.31 feedback



This material is based upon work supported by the National Science Foundation (NSF) under Grant PHY-0555655 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.

Copyright © 2024 Bruce Allen for the LIGO Scientific Collaboration