World Community Grid - View Thread - Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

World Community Grid Forums

Category: Support

Forum: GPU Support Forum

Thread: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

Quick Go »

No member browsing this thread

Thread Status: Active
Total posts in this thread: 11

[ ]

Author

This topic has been viewed 7712 times and has 10 replies

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

The list of GPU boards that are excluded from running HCC GPU is: https://secure.worldcommunitygrid.org/help/viewTopic.do?shortName=GPU#610
If your GPU board is not on this list and sometimes produces valid results but sometimes fails then please post on this thread and explain the errors and their frequency. At least one post has already been made. Obvious possibilities are over-clocking, hardware problems, drivers, settings, general software environment, motherboard problems, etc. And perhaps there are some GPU board architectures out there that will always act this way. So please tell us a lot in these posts.

Lawrence

[Oct 12, 2012 4:13:40 AM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

GTX 470, windows 7 x64, nvidia 301.42, boinc 7.0.28 x64

10/12/2012 1:45:46 AM | World Community Grid | Starting task X0930069530467200606141931_1 using hcc1 version 656 (nvidia_hcc1) in slot 4
10/12/2012 1:46:07 AM | World Community Grid | Computation for task X0930069530467200606141931_1 finished
10/12/2012 1:46:07 AM | World Community Grid | Output file X0930069530467200606141931_1_0 for task X0930069530467200606141931_1 absent
10/12/2012 1:46:07 AM | World Community Grid | Starting task X0930069530469200606141931_1 using hcc1 version 656 (nvidia_hcc1) in slot 4
10/12/2012 1:46:29 AM | World Community Grid | Computation for task X0930069530469200606141931_1 finished
10/12/2012 1:46:29 AM | World Community Grid | Output file X0930069530469200606141931_1_0 for task X0930069530469200606141931_1 absent
10/12/2012 1:46:29 AM | World Community Grid | Starting task X0930069530485200606141931_0 using hcc1 version 656 (nvidia_hcc1) in slot 4
10/12/2012 1:46:51 AM | World Community Grid | Computation for task X0930069530485200606141931_0 finished
10/12/2012 1:46:51 AM | World Community Grid | Output file X0930069530485200606141931_0_0 for task X0930069530485200606141931_0 absent

edit: nevermind, I had SLI enabled and forgot :(

----------------------------------------
[Edit 1 times, last edit by Former Member at Oct 12, 2012 6:53:46 AM]

[Oct 12, 2012 6:49:16 AM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

I started processing GPU tasks only today with my GeForce GT 430.

It seems to work but, like others, I set it to run only when idle because of the refresh rate issue.

That said, only one of my gpu tasks has been validated yet and two report an error on the Result Status page. Two others are pending validation.

Also, in the boinc manager one (gpu) task is stuck at 100% but the elapsed time is only 2 seconds and the status is "Computation error". The following messages may be related:

12-Oct-2012 9:41:31 PM World Community Grid Computation for task X0960069070324200605301847_1 finished
12-Oct-2012 9:41:31 PM World Community Grid Output file X0960069070324200605301847_1_0 for task X0960069070324200605301847_1 absent

I'll wait and see but 1 out of 3 doesn't look very good... It is fast though. ;)

OS: Win7pro 64-bit
BOINC version: 6.10.58 (official from WCG)
GPU: nvidia GeForce GT 430
Display driver: 275.33 (maybe I should update this)
CPU: Intel i5-2500 Sandy Bridge (monitor plugged into discrete card)
***NO CPU/GPU OVERCLOCKING(except Intel turbo boost)

[Oct 13, 2012 3:53:45 AM]

Tomahawk4196
Advanced Cruncher
USA
Joined: Aug 16, 2007
Post Count: 93
Status: Offline
Project Badges:

5 year badge for Human Proteome Folding - Phase 2

180 day badge for Discovering Dengue Drugs - Together

90 day badge for Nutritious Rice for the World

2 year badge for Help Fight Childhood Cancer

2 year badge for Help Cure Muscular Dystrophy - Phase 2

14 day badge for Discovering Dengue Drugs - Together - Phase 2

5 year badge for The Clean Energy Project - Phase 2

180 day badge for Computing for Clean Water

2 year badge for Drug Search for Leishmaniasis

2 year badge for GO Fight Against Malaria

2 year badge for Computing for Sustainable Water

50 year badge for Mapping Cancer Markers

2 year badge for Uncovering Genome Mysteries

2 year badge for Outsmart Ebola Together

2 year badge for FightAIDS@Home - Phase 2

50 year badge for Smash Childhood Cancer

2 year badge for Microbiome Immunity Project

180 day badge for Africa Rainfall Project

5 year badge for OpenPandemics - COVID-19


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

nVidia GeForce GTX 460, stock clock, nVidia driver 296.10
Win7pro 64-bit, CPU i7-2600K

Most HCC units are failing, tho some appear to be OK.
Fail error:

Result Name: X0930069081519200605101155_ 0--

<core_client_version>7.0.28</core_client_version>
<![CDATA[
<message>
The pipe is being closed. (0xe8) - exit code 232 (0xe8)
</message>
<stderr_txt>
Commandline: projects/www.worldcommunitygrid.org/wcg_hcc1_img_6.56_windows_intelx86__nvidia_hcc1 X0930069081519200605101155.jp2 --device 0
<app_init_data>
<major_version>7</major_version>
<minor_version>0</minor_version>
<release>28</release>
<app_version>656</app_version>
<app_name>hcc1</app_name>
<project_preferences>

<color_scheme>Tahiti Sunset</color_scheme>
<max_frames_sec>7</max_frames_sec>
<max_gfx_cpu_pct>5.0</max_gfx_cpu_pct>
</project_preferences>

<project_dir>C:\ProgramData\BOINC/projects/www.worldcommunitygrid.org</project_dir>
<boinc_dir>C:\ProgramData\BOINC</boinc_dir>
<wu_name>X0930069081519200605101155</wu_name>
<result_name>X0930069081519200605101155_0</result_name>
<comm_obj_name>boinc_4</comm_obj_name>
<slot>6</slot>
<wu_cpu_time>0.000000</wu_cpu_time>
<starting_elapsed_time>0.000000</starting_elapsed_time>
<using_sandbox>0</using_sandbox>
<user_total_credit>5798837.274021</user_total_credit>
<user_expavg_credit>8094.784080</user_expavg_credit>
<host_total_credit>1526397.264985</host_total_credit>
<host_expavg_credit>5163.523442</host_expavg_credit>
<resource_share_fraction>1.000000</resource_share_fraction>
<checkpoint_period>60.000000</checkpoint_period>
<fraction_done_start>0.000000</fraction_done_start>
<fraction_done_end>1.000000</fraction_done_end>
<gpu_type>NVIDIA</gpu_type>
<gpu_device_num>0</gpu_device_num>
<gpu_opencl_dev_index>0</gpu_opencl_dev_index>
<ncpus>1.000000</ncpus>
<rsc_fpops_est>13685181562850.000000</rsc_fpops_est>
<rsc_fpops_bound>273703631257000.000000</rsc_fpops_bound>
<rsc_memory_bound>78643200.000000</rsc_memory_bound>
<rsc_disk_bound>50000000.000000</rsc_disk_bound>
<computation_deadline>1350661930.000000</computation_deadline>
<vbox_window>0</vbox_window>
</app_init_data>
INFO: gpu_type set in init_data.xml to NVIDIA
INFO: gpu_device_num set in init_data.xml to 0
Boinc requested NVIDIA gpu device number 0
ERROR: .\VerifyGPU.cpp:65 Unknown
19:12:46 (4460): called boinc_finish

</stderr_txt>
]]>

----------------------------------------

[Oct 13, 2012 4:11:36 AM]

Poulpe
Cruncher
Joined: Nov 22, 2005
Post Count: 6
Status: Offline
Project Badges:

2 year badge for Human Proteome Folding - Phase 2

90 day badge for Discovering Dengue Drugs - Together

1 year badge for Help Fight Childhood Cancer

14 day badge for Influenza Antiviral Drug Search

1 year badge for Help Cure Muscular Dystrophy - Phase 2

1 year badge for Computing for Clean Water

1 year badge for Drug Search for Leishmaniasis

1 year badge for GO Fight Against Malaria

14 day badge for Computing for Sustainable Water

45 day badge for Uncovering Genome Mysteries

1 year badge for Outsmart Ebola Together

1 year badge for FightAIDS@Home - Phase 2

1 year badge for Microbiome Immunity Project


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

Nvidia GT140

14/10/2012 22:09:33 | World Community Grid | Output file X0900069220870200606071744_0_0 for task X0900069220870200606071744_0 absent
14/10/2012 22:09:33 | World Community Grid | Starting task X0900069220863200606071745_0 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:34 | World Community Grid | Finished download of X0900069220872200606071744_X0900069220872200606071744.jp2
14/10/2012 22:09:34 | World Community Grid | Finished download of X0900069220858200606071745_X0900069220858200606071745.jp2
14/10/2012 22:09:34 | World Community Grid | Started download of X0900069220841200606071744_X0900069220841200606071744.jp2
14/10/2012 22:09:36 | World Community Grid | Computation for task X0900069220863200606071745_0 finished
14/10/2012 22:09:36 | World Community Grid | Output file X0900069220863200606071745_0_0 for task X0900069220863200606071745_0 absent
14/10/2012 22:09:36 | World Community Grid | Starting task X0900069220872200606071744_0 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:37 | World Community Grid | Finished download of X0900069220841200606071744_X0900069220841200606071744.jp2
14/10/2012 22:09:39 | World Community Grid | Computation for task X0900069220872200606071744_0 finished
14/10/2012 22:09:39 | World Community Grid | Output file X0900069220872200606071744_0_0 for task X0900069220872200606071744_0 absent
14/10/2012 22:09:39 | World Community Grid | Starting task X0900069220858200606071745_0 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:42 | World Community Grid | Computation for task X0900069220858200606071745_0 finished
14/10/2012 22:09:42 | World Community Grid | Output file X0900069220858200606071745_0_0 for task X0900069220858200606071745_0 absent
14/10/2012 22:09:42 | World Community Grid | Starting task X0900069220866200606071744_1 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:45 | World Community Grid | Computation for task X0900069220866200606071744_1 finished
14/10/2012 22:09:45 | World Community Grid | Output file X0900069220866200606071744_1_0 for task X0900069220866200606071744_1 absent
14/10/2012 22:09:45 | World Community Grid | Starting task X0900069220868200606071744_0 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:48 | World Community Grid | Computation for task X0900069220868200606071744_0 finished
14/10/2012 22:09:48 | World Community Grid | Output file X0900069220868200606071744_0_0 for task X0900069220868200606071744_0 absent
14/10/2012 22:09:48 | World Community Grid | Starting task X0900069220869200606071744_1 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:51 | World Community Grid | Computation for task X0900069220869200606071744_1 finished
14/10/2012 22:09:51 | World Community Grid | Output file X0900069220869200606071744_1_0 for task X0900069220869200606071744_1 absent
14/10/2012 22:09:51 | World Community Grid | Starting task X0900069220852200606071744_1 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:54 | World Community Grid | Computation for task X0900069220852200606071744_1 finished
14/10/2012 22:09:54 | World Community Grid | Output file X0900069220852200606071744_1_0 for task X0900069220852200606071744_1 absent
14/10/2012 22:09:54 | World Community Grid | Starting task X0900069220841200606071744_0 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:09:57 | World Community Grid | Computation for task X0900069220841200606071744_0 finished
14/10/2012 22:09:57 | World Community Grid | Output file X0900069220841200606071744_0_0 for task X0900069220841200606071744_0 absent
14/10/2012 22:09:57 | World Community Grid | Starting task X0900069220865200606071744_0 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:10:00 | World Community Grid | Computation for task X0900069220865200606071744_0 finished
14/10/2012 22:10:00 | World Community Grid | Output file X0900069220865200606071744_0_0 for task X0900069220865200606071744_0 absent
14/10/2012 22:10:00 | World Community Grid | Starting task X0900069220871200606071744_1 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:10:04 | World Community Grid | Computation for task X0900069220871200606071744_1 finished
14/10/2012 22:10:04 | World Community Grid | Output file X0900069220871200606071744_1_0 for task X0900069220871200606071744_1 absent
14/10/2012 22:10:04 | World Community Grid | Starting task X0900069220854200606071744_1 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:10:07 | World Community Grid | Computation for task X0900069220854200606071744_1 finished
14/10/2012 22:10:07 | World Community Grid | Output file X0900069220854200606071744_1_0 for task X0900069220854200606071744_1 absent
14/10/2012 22:10:07 | World Community Grid | Starting task X0900069220850200606071744_1 using hcc1 version 656 (nvidia_hcc1) in slot 6
14/10/2012 22:10:10 | World Community Grid | Computation for task X0900069220850200606071744_1 finished
14/10/2012 22:10:10 | World Community Grid | Output file X0900069220850200606071744_1_0 for task X0900069220850200606071744_1 absent

[Oct 14, 2012 9:46:18 PM]

Tomahawk4196
Advanced Cruncher
USA
Joined: Aug 16, 2007
Post Count: 93
Status: Offline
Project Badges:


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

I read somewhere in these forums that if you turn off hyperthreading, the GPU gets a dedicated CPU core, and the 'computation error' goes away. I'm trying it now, and after about 30 WUs, it is having no problems ... fingers crossed.

Edit: After about 6 hours of running, I got at least 50 examples of "computation error". I'm shutting down GPU on my machines, please e-mail me when all these bugs are sorted out.

----------------------------------------

----------------------------------------
[Edit 1 times, last edit by Tomahawk4196 at Oct 17, 2012 10:55:15 AM]

[Oct 17, 2012 12:11:23 AM]

armstrdj
Former World Community Grid Tech
Joined: Oct 21, 2004
Post Count: 695
Status: Offline
Project Badges:

14 day badge for Help Cure Muscular Dystrophy

90 day badge for The Clean Energy Project

90 day badge for Influenza Antiviral Drug Search

2 year badge for Discovering Dengue Drugs - Together - Phase 2

2 year badge for The Clean Energy Project - Phase 2

2 year badge for Computing for Clean Water

10 year badge for Mapping Cancer Markers

2 year badge for Africa Rainfall Project

2 year badge for OpenPandemics - COVID-19


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

Tomahawk, the error you are seeing is coming on the first OpenCl call that the code tries to make. It could be that your OpenCL ICD is corrupt. I would check to make sure you have the latest driver installed and if you do reinstall. Also if you ever use Remote Desktop Connection to connect to the machine where you are seeing the errors this could be causing problems.

Thanks,
armstrdj

[Oct 17, 2012 6:18:15 PM]

nanoprobe
Master Cruncher
Classified
Joined: Aug 29, 2008
Post Count: 2998
Status: Offline
Project Badges:

10 year badge for Help Fight Childhood Cancer

5 year badge for Help Cure Muscular Dystrophy - Phase 2

20 year badge for Computing for Clean Water

5 year badge for Drug Search for Leishmaniasis

5 year badge for GO Fight Against Malaria

5 year badge for Computing for Sustainable Water

20 year badge for Mapping Cancer Markers

20 year badge for Uncovering Genome Mysteries

50 year badge for Outsmart Ebola Together

10 year badge for FightAIDS@Home - Phase 2

20 year badge for Smash Childhood Cancer

20 year badge for Microbiome Immunity Project

10 year badge for Africa Rainfall Project

20 year badge for OpenPandemics - COVID-19


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

nVidia GeForce GTX 460, stock clock, nVidia driver 296.10

<--- Update your drivers. This is a know problem driver for GPU computing.

----------------------------------------

In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.

[Oct 17, 2012 7:30:41 PM]

Tomahawk4196
Advanced Cruncher
USA
Joined: Aug 16, 2007
Post Count: 93
Status: Offline
Project Badges:


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

Thanks, that did the trick. Updated driver to 306.97, all things are now running smoothly.

Any word on a dedicated GPU page?

Thanks again

----------------------------------------

----------------------------------------
[Edit 1 times, last edit by Tomahawk4196 at Oct 19, 2012 11:27:48 PM]

[Oct 19, 2012 11:27:28 PM]

pmm1018
Senior Cruncher
USA
Joined: Dec 29, 2006
Post Count: 222
Status: Offline
Project Badges:

180 day badge for Help Fight Childhood Cancer

20 year badge for The Clean Energy Project - Phase 2

20 year badge for Outsmart Ebola Together

20 year badge for FightAIDS@Home - Phase 2


Re: Please Post if GPU Board OCCASIONALLY fails and is not on excluded list

I have a NVIDIA Quadro 1000M GPU running in tandem with an Intel Core i7-2860QM CPU. I was having overheat issues of the GPU with my Dell Precision laptop. It would work fine on about a dozen nvidia_hcc1 work units and then would crap out. I solved this problem (for now) by reducing the "use at most" CPU time from 100% down to 75% in the BOINC processor usage tab under Tools - Computing preferences.

----------------------------------------

[Oct 21, 2012 5:37:50 PM]

[ ]