World Community Grid - View Thread - GPU Work Units - Post Your Tech Support Questions Here

World Community Grid Forums

Category: Support

Forum: GPU Support Forum

Thread: GPU Work Units - Post Your Tech Support Questions Here

Quick Go »

No member browsing this thread

Thread Status: Active
Thread Type: Sticky Thread
Total posts in this thread: 290

[ ]

Author

This topic has been viewed 555910 times and has 289 replies

[AF>Libristes]Maeda
Cruncher
Joined: Sep 1, 2011
Post Count: 43
Status: Offline
Project Badges:

14 day badge for Human Proteome Folding - Phase 2

45 day badge for The Clean Energy Project - Phase 2

14 day badge for Drug Search for Leishmaniasis

14 day badge for GO Fight Against Malaria

20 year badge for Mapping Cancer Markers

45 day badge for Uncovering Genome Mysteries

180 day badge for Outsmart Ebola Together

180 day badge for FightAIDS@Home - Phase 2

2 year badge for Microbiome Immunity Project

1 year badge for Africa Rainfall Project

1 year badge for OpenPandemics - COVID-19


Re: GPU Work Units - Post Your Tech Support Questions Here

Hello,
OPNG beta WU was working well, but now I am getting this error message for my Intel iGPU HD4600 (on Linux):


Result Name: OPNG_ 0062351_ 00010_ 0--
<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
process exited with code 10 (0xa, -246)</message>
<stderr_txt>
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_intel_gpu_102 -jobs OPNG_0062351_00010.job -input OPNG_0062351_00010.zip -seed 643372163 -wcgruns 2100 -wcgdpf 42
INFO: Using gpu device from app init data 0
INFO:[20:18:16] Start AutoGrid...

autogrid4: Successful Completion.
INFO:[20:19:54] End AutoGrid...
INFO:[20:19:54] Start AutoDock for OB3ZINC000113508841_2--7jji_001--ALYS417_inert.dpf(Job #0)...
/dev/dri/card0 not authenticated
/dev/dri/card0 not authenticated
beignet-opencl-icd: no supported GPU found, this is probably the wrong opencl-icd package for this hardware
(If you have multiple ICDs installed and OpenCL works, you can ignore this message)
GPU not found: type=intel_gpu, opencl_device_index=0, device_num=0
Error: boinc_get_opencl_ids() failed with error -1
20:19:54 (26116): called boinc_finish(10)

</stderr_txt>
]]>

Are other people using that iGPU having same issue?

Here are packages relative to opencl which are installed:


sudo apt list --installed *opencl*

beignet-opencl-icd/stable,now 1.3.2-6+b10 amd64
boinc-client-opencl/stable,now 7.14.2+dfsg-3 amd64
ocl-icd-libopencl1/stable,now 2.2.12-2 amd64

Thanks.

[Jul 15, 2021 6:00:10 AM]

Wisesooth
Cruncher
United States
Joined: Aug 5, 2016
Post Count: 9
Status: Offline
Project Badges:

5 year badge for Outsmart Ebola Together

2 year badge for FightAIDS@Home - Phase 2

5 year badge for Microbiome Immunity Project

2 year badge for Africa Rainfall Project

2 year badge for OpenPandemics - COVID-19


Re: GPU Work Units - Post Your Tech Support Questions Here

The open pandemics project launched a GPU task using an Intel GPU. It was suspending indefinitely stating that the Intel GPU was already in use. Had to abort the task. Apparently, this is a problem with the latest release of BOINC. Saw a lot of bug comments on the BOINC forum about GPU issues. Thought you might like to know.

[Jul 19, 2021 9:27:59 PM]

HGRAY515
Cruncher
Joined: Jun 23, 2011
Post Count: 11
Status: Offline
Project Badges:

180 day badge for Human Proteome Folding - Phase 2

90 day badge for Help Fight Childhood Cancer

14 day badge for Help Cure Muscular Dystrophy - Phase 2

180 day badge for The Clean Energy Project - Phase 2

45 day badge for Outsmart Ebola Together


Re: GPU Work Units - Post Your Tech Support Questions Here

I am sure this may have been answered before….if so, sorry. I know not all WCG projects are viable for developing a GPU option. Are there long running projects (like maybe MCM) on WCG that would benefit from adding a GPU Option?. If so, is the roadblock funding ($$$) to develop and test? If funding is the issue, could we attempt a crowd funding to push that along?

[Jul 22, 2021 6:25:01 PM]

Grumpy Swede
Master Cruncher
Svíþjóð
Joined: Apr 10, 2020
Post Count: 2094
Status: Offline
Project Badges:

14 day badge for FightAIDS@Home - Phase 2

1 year badge for Microbiome Immunity Project

90 day badge for Africa Rainfall Project


Re: GPU Work Units - Post Your Tech Support Questions Here

Hehe, a GPU WU, containing just one "job". Never seen that before.

<core_client_version>7.16.7</core_client_version>
<![CDATA[
<stderr_txt>
projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_windows_x86_64__opencl_nvidia_102 -jobs OPNG_0075767_00221.job -input OPNG_0075767_00221.zip -seed 1694001489 -wcgruns 50 -wcgdpf 1
INFO: Using gpu device from app init data 0
INFO:[14:02:38] Start AutoGrid...

autogrid4: Successful Completion.
INFO:[14:03:47] End AutoGrid...
INFO:[14:03:47] Start AutoDock for OB3ZINC001193226783--7jji_001--ALYS417_inert.dpf(Job #0)...
OpenCL device: GeForce GTX 980
INFO:[14:03:52] End AutoDock...
INFO:Cpu time = 74.546875
14:03:53 (3136): called boinc_finish(0)

</stderr_txt>
]]>

[Aug 14, 2021 12:14:24 PM]

fatespawn
Cruncher
Joined: Jan 17, 2017
Post Count: 5
Status: Offline
Project Badges:

20 year badge for OpenPandemics - COVID-19


Re: GPU Work Units - Post Your Tech Support Questions Here

Hi all,

I just upgraded my Ubuntu rig with an RX560. It appears I'm getting work units sent to me, but I don't see any in the tasks nor in the results log.

Where should I start?

Thanks.

Tue 17 Aug 2021 01:43:52 PM CDT | World Community Grid | Requesting new tasks for CPU and AMD/ATI GPU
Tue 17 Aug 2021 01:43:53 PM CDT | World Community Grid | Scheduler request completed: got 4 new tasks
Tue 17 Aug 2021 01:43:53 PM CDT | World Community Grid | Project requested delay of 121 seconds

[Aug 17, 2021 6:59:22 PM]

Acibant
Advanced Cruncher
USA
Joined: Apr 15, 2020
Post Count: 126
Status: Offline
Project Badges:

50 year badge for Mapping Cancer Markers

10 year badge for Africa Rainfall Project

5 year badge for OpenPandemics - COVID-19


Re: GPU Work Units - Post Your Tech Support Questions Here

GPU work units can be hard to get ahold of. It's clear from the portion of the log you pasted that the client is checking for them. But are you operating under the assumption that those four tasks were GPU work units? They may very well have been CPU work units. Can you check your results status page for four units sent around 18:43? (The times on that page are shown in UTC.) If they start with OPNG then we'll know they were GPU work units.

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

[Aug 17, 2021 10:35:18 PM]

fatespawn
Cruncher
Joined: Jan 17, 2017
Post Count: 5
Status: Offline
Project Badges:


Re: GPU Work Units - Post Your Tech Support Questions Here

Thanks, yes, my misunderstanding on the log file. I checked the work and no OPNG yet for my Ubuntu rig. I have a MacPro too with an rx480 that never gets work units either… however just yesterday a different macpro with a newly installed rx580 just got some work units which was a surprise to me (since I had never seen them on my 480 before).

I guess I’ll start watching “my contribution” more closely.

[Aug 19, 2021 11:17:06 PM]

fatespawn
Cruncher
Joined: Jan 17, 2017
Post Count: 5
Status: Offline
Project Badges:


Re: GPU Work Units - Post Your Tech Support Questions Here

So, my Ubuntu rig is definitely getting OPNG work but it keeps returning "Error". Is there something I can do?

Result Log

Result Name: OPNG_ 0078228_ 00119_ 0--

<core_client_version>7.16.6</core_client_version>
<![CDATA[
<message>
process exited with code 193 (0xc1, -63)</message>
<stderr_txt>
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102 -jobs OPNG_0078228_00119.job -input OPNG_0078228_00119.zip -seed 19128469 -wcgruns 2500 -wcgdpf 50
INFO: Using gpu device from app init data 0
INFO:[00:11:59] Start AutoGrid...

autogrid4: Successful Completion.
INFO:[00:12:40] End AutoGrid...
INFO:[00:12:40] Start AutoDock for OB3ZINC000242503608--7jji_001--ALYS417_inert.dpf(Job #0)...
SIGSEGV: segmentation violation
Stack trace (22 frames):
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x4532b2]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x153c0)[0x7f83806583c0]
/lib/x86_64-linux-gnu/libc.so.6(cfree+0x20)[0x7f83804ec870]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libamdocl64.so(+0xced22)[0x7f837a6aad22]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libamdocl64.so(+0xd0429)[0x7f837a6ac429]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libamdocl64.so(+0x81b63)[0x7f837a65db63]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libamdocl64.so(+0x8ba7e)[0x7f837a667a7e]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libamdocl64.so(+0x7b8b5)[0x7f837a6578b5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x1247f)[0x7f838065547f]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libamdocl64.so(clIcdGetPlatformIDsKHR+0xad)[0x7f837a6579dd]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libOpenCL.so.1(+0x3565)[0x7f83809d1565]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libOpenCL.so.1(+0x564e)[0x7f83809d364e]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x1247f)[0x7f838065547f]
/opt/amdgpu-pro/lib/x86_64-linux-gnu/libOpenCL.so.1(clGetPlatformIDs+0x15)[0x7f83809d1be5]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x46ac50]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x46b087]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x42e65c]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x4238c5]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x41dba9]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x407e20]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf3)[0x7f83804760b3]
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_ati_102[0x408662]

Exiting...

</stderr_txt>
]]>

[Aug 22, 2021 5:57:48 PM]

Acibant
Advanced Cruncher
USA
Joined: Apr 15, 2020
Post Count: 126
Status: Offline
Project Badges:


Re: GPU Work Units - Post Your Tech Support Questions Here

From an FAQ entry it seems that you may be able to correct this by exiting and restarting BOINC (or restarting the system). If it persists after that then there may be an issue with your memory that would require testing using this guide for your Ubuntu system to prove out.

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

[Aug 22, 2021 6:50:30 PM]

Collin53
Cruncher
Joined: Apr 29, 2007
Post Count: 1
Status: Offline


Re: GPU Work Units - Post Your Tech Support Questions Here

I have a Gigabyte NVIDIA GeForce GTX680 that when it runs the GPU work unit will reset it's display resolution settings to 640x480 when it's normally at 2560x1440 on an ASUS PB278 monitor, most annoying.

Is anyone else having this issue?

Using the latest NVIDIA Game Driver 30.0.14.7168 on the latest Windows 10 Pro with all known updates applied.
Edition Windows 10 Pro
Version 21H1
Installed on ‎16/‎08/‎2020
OS build 19043.1165
Experience Windows Feature Experience Pack 120.2212.3530.0
Processor Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz 3.90 GHz
Installed RAM 32.0 GB
System type 64-bit operating system, x64-based processor

Thanks

[Aug 31, 2021 3:40:53 AM]

[ ]