ATI OpenCL MultiBeam 6.10 problem..

Message boards : Number crunching : ATI OpenCL MultiBeam 6.10 problem..
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 4 · 5 · 6 · 7

AuthorMessage
Profile Sunny129
Avatar

Send message
Joined: 7 Nov 00
Posts: 190
Credit: 3,163,755
RAC: 0
United States
Message 1087600 - Posted: 17 Mar 2011, 0:09:08 UTC

ok guys,

using BOINC v6.10.58, i tried Claggy's suggestion and edited the <cmdline>-period_iterations_num 2 -instances_per_device 1</cmdline> line in SETI's app_info.xml file...that is, i changed instances per device from 1 to 2.

1) on the first attempt, there was no cc-Config.xml file in the BOINC data directory either disabling or enabling GPUs. no success - the tasks behaved the same as they did with scenario #1 under v6.10.58. here is the task that errored out under this configuration: http://setiathome.berkeley.edu/result.php?resultid=1836692668.
2) i tried it again, this time with a "<use_all_gpus>1</use_all_gpus>" cc-config.xml file, and i got the same results. here is the task that errored out under this configuration: http://setiathome.berkeley.edu/result.php?resultid=1836692667.
3) i also tried it with a "<use_all_gpus>0</use_all_gpus>" cc-config.xml file - again, same results. here is the task that errored out under this configuration: http://setiathome.berkeley.edu/result.php?resultid=1836692684.

i'm gonna give it another try using BOINC v6.12.18 later tonight. i'll report back...
ID: 1087600 · Report as offensive
Profile Sunny129
Avatar

Send message
Joined: 7 Nov 00
Posts: 190
Credit: 3,163,755
RAC: 0
United States
Message 1087627 - Posted: 17 Mar 2011, 2:32:43 UTC

ok,

using BOINC v6.12.18, i performed the same experiment as above.

1) on the first attempt, there was no cc-Config.xml file in the BOINC data directory either disabling or enabling GPUs. no success - the tasks behaved the same as they did with scenario #1 under v6.10.58 (immediate errors). here is the task that errored out under this configuration: http://setiathome.berkeley.edu/result.php?resultid=1836692655.
2) i tried it again, this time with a "<use_all_gpus>1</use_all_gpus>" cc-config.xml file, and i got the same results (immediate errors). here is the task that errored out under this configuration: http://setiathome.berkeley.edu/result.php?resultid=1836692620.
3) i also tried it with a "<use_all_gpus>0</use_all_gpus>" cc-config.xml file - again, same results (this time they were subsequent errors). here is the task that errored out under this configuration: http://setiathome.berkeley.edu/result.php?resultid=1836692652.

...so unfortunately changing "instances per device" from 1 to 2 in SETI's app_info.xml file did not change anything.
ID: 1087627 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6324
Credit: 106,370,077
RAC: 121
Russia
Message 1087700 - Posted: 17 Mar 2011, 5:59:40 UTC - in response to Message 1087627.  

Now we sure it will not work at least :)
The reason is the same - app tries to pick up non-existent device.
ID: 1087700 · Report as offensive
Previous · 1 . . . 4 · 5 · 6 · 7

Message boards : Number crunching : ATI OpenCL MultiBeam 6.10 problem..


 
©2022 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.