Advanced search

Message boards : Graphics cards (GPUs) : Work units fail on every run with a 9600GT

Author Message
nickth
Send message
Joined: 11 May 08
Posts: 2
Credit: 93,529
RAC: 0
Level

Scientific publications
wat
Message 10442 - Posted: 7 Jun 2009 | 14:35:08 UTC

Every time i try to run GPUGrid on my computer which is a

Intel Quad core Q8200 with 3GB of ram and a 9600GT driver 185.85 not overclocked running Windows 7

This is the out put file from the task

<core_client_version>6.6.31</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
# Using CUDA device 0
# Device 0: "GeForce 9600 GT"
# Clock rate: 1625000 kilohertz
# Total amount of global memory: 536870912 bytes
# Number of multiprocessors: 8
# Number of cores: 64
# Amber: readparm : Reading parm file parameters
# PARM file in AMBER 7 format
# Encounter 10-12 H-bond term
WARNING: parameters.cu, line 568: Found zero 10-12 H-bond term.
WARNING: parameters.cu, line 568: Found zero 10-12 H-bond term.
MDIO ERROR: cannot open file "restart.coor"
Cuda error: Kernel [frc_sum_kernel_angle] failed in file 'force.cu' in line 298 : unknown error.

</stderr_txt>
]]>

Its the same every time i try wether the task last 30 sec or 1 hour 25 min they all error out

Hope you can help

Profile Hydropower
Avatar
Send message
Joined: 3 Apr 09
Posts: 70
Credit: 6,003,024
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 10443 - Posted: 7 Jun 2009 | 14:55:25 UTC

The thread below deals with the same (yet unresolved) topic:
http://www.gpugrid.net/forum_thread.php?id=1130
____________
Join team Bletchley Park, the innovators.

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 10509 - Posted: 12 Jun 2009 | 21:07:11 UTC

nickth, might help if your computers were not hidden. Otherwise.. is the problem still there? Did you try any of the usual suggestions in such cases, i.e. run 3D Mark or FurMark for an hour, check temps, downclock etc.

MrS
____________
Scanning for our furry friends since Jan 2002

nickth
Send message
Joined: 11 May 08
Posts: 2
Credit: 93,529
RAC: 0
Level

Scientific publications
wat
Message 10569 - Posted: 14 Jun 2009 | 14:58:45 UTC - in response to Message 10509.

Yes i managed to get it to work

<core_client_version>6.6.36</core_client_version>
<![CDATA[
<stderr_txt>
# Using CUDA device 0
# Device 0: "GeForce 9600 GT"
# Clock rate: 1625000 kilohertz
# Total amount of global memory: 536870912 bytes
# Number of multiprocessors: 8
# Number of cores: 64
# Amber: readparm : Reading parm file parameters
# PARM file in AMBER 7 format
# Encounter 10-12 H-bond term
WARNING: parameters.cu, line 568: Found zero 10-12 H-bond term.
WARNING: parameters.cu, line 568: Found zero 10-12 H-bond term.
MDIO ERROR: cannot open file "restart.coor"
# Time per step: 240.071 ms
# Approximate elapsed time for entire WU: 72021.185 s
called boinc_finish

</stderr_txt>
]]>

The only way to get it to work was to let one core be used for the gpu

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 10570 - Posted: 14 Jun 2009 | 15:09:23 UTC - in response to Message 10569.

That's very unusual. What did you change, a cc_config option or a BOINC setting? Would you mind verifying that this was indeed the cure but switching back and reporting if tasks fail again?

MrS
____________
Scanning for our furry friends since Jan 2002

Post to thread

Message boards : Graphics cards (GPUs) : Work units fail on every run with a 9600GT

//