Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Thread Type: Sticky Thread
Total posts in this thread: 290
Posts: 290   Pages: 29   [ Previous Page | 6 7 8 9 10 11 12 13 14 15 | Next Page ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 580183 times and has 289 replies Next Thread
Dfirebug
Cruncher
Joined: Jan 6, 2021
Post Count: 5
Status: Offline
Reply to this Post  Reply with Quote 
I am having a problem with my GPU not always using its full power and the usage is not constant



this is what my GPU usage is doing I have an Nvidia RTX 2080 super and a ryzen 5 1400 CPU a pny 240 GB SSD that can consistently go 500 MB/s I don't know what is wrong and I would like to use my GPU's full power my usage is staying at about 0-5% until suddenly it spikes to 100% for about a second and then back to 0%
----------------------------------------
[Edit 1 times, last edit by Dfirebug at Apr 27, 2021 5:07:33 PM]
[Apr 27, 2021 5:06:16 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Falconet
Master Cruncher
Portugal
Joined: Mar 9, 2009
Post Count: 3295
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: I am having a problem with my GPU not always using its full power and the usage is not constant



this is what my GPU usage is doing I have an Nvidia RTX 2080 super and a ryzen 5 1400 CPU a pny 240 GB SSD that can consistently go 500 MB/s I don't know what is wrong and I would like to use my GPU's full power my usage is staying at about 0-5% until suddenly it spikes to 100% for about a second and then back to 0%



Completely normal behaviour. Each spike is basically the GPU working on 1 job. Each GPU task you receive likely has at a couple dozen jobs.

If you want to increase the GPU usage, you could create an app_config file.
----------------------------------------


AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W
AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W
AMD Ryzen 7 7730U 8C/16T 3.0 GHz
[Apr 27, 2021 5:58:49 PM]   Link   Report threatening or abusive post: please login first  Go to top 
William Albert
Cruncher
Joined: Apr 5, 2020
Post Count: 36
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: I am having a problem with my GPU not always using its full power and the usage is not constant

During the stress test yesterday, several Nvidia WUs on two of my machines with Nvidia GeForce GT 720 cards errored out. As far as I can tell, they were running without any problems, but were terminated because the WU has a hard cap on the elapsed time limit.

It's not clear why such a strict cap is in place when the WU deadline is days away, especially when GPUs can vary wildly (much more than CPUs) in terms of throughput.

Unless lower-end GPUs are simply unwanted by the project, these time caps may need to be revisited.

Details below:

Apr 27 07:02:45 redacted-host-1 boinc[3490]: 27-Apr-2021 00:02:45 [World Community Grid] Aborting task OPNG_0005115_00189_0: exceeded elapsed time limit 7973.34 (628994.24G/83.94G)
Apr 27 11:01:53 redacted-host-1 boinc[3490]: 27-Apr-2021 04:01:53 [World Community Grid] Aborting task OPNG_0005115_00438_0: exceeded elapsed time limit 7142.37 (628994.24G/88.07G)
Apr 27 13:00:58 redacted-host-1 boinc[3490]: 27-Apr-2021 06:00:58 [World Community Grid] Aborting task OPNG_0005115_00455_0: exceeded elapsed time limit 7142.37 (628994.24G/86.90G)
Apr 27 16:51:02 redacted-host-1 boinc[6845]: 27-Apr-2021 09:51:02 [World Community Grid] Aborting task OPNG_0005722_00254_1: exceeded elapsed time limit 7320.54 (628994.24G/85.63G)


Apr 27 08:17:17 redacted-host-2 boinc[840]: 27-Apr-2021 01:17:17 [World Community Grid] Aborting task OPNG_0005909_00256_0: exceeded elapsed time limit 7313.30 (628994.24G/87.38G)
Apr 27 10:17:19 redacted-host-2 boinc[840]: 27-Apr-2021 03:17:19 [World Community Grid] Aborting task OPNG_0005909_00165_0: exceeded elapsed time limit 7198.67 (628994.24G/87.38G)
Apr 27 12:17:21 redacted-host-2 boinc[840]: 27-Apr-2021 05:17:21 [World Community Grid] Aborting task OPNG_0005668_00052_0: exceeded elapsed time limit 7198.67 (628994.24G/87.47G)
Apr 27 14:17:16 redacted-host-2 boinc[840]: 27-Apr-2021 07:17:16 [World Community Grid] Aborting task OPNG_0005668_00061_0: exceeded elapsed time limit 7190.93 (628994.24G/87.47G)
Apr 27 16:17:09 redacted-host-2 boinc[840]: 27-Apr-2021 09:17:09 [World Community Grid] Aborting task OPNG_0005971_00151_0: exceeded elapsed time limit 7190.93 (628994.24G/85.43G)

Spot-checking the WU results shows the following:
<core_client_version>7.16.6</core_client_version>
<![CDATA[
<message>
exceeded elapsed time limit 7973.34 (628994.24G/83.94G)</message>
<stderr_txt>
../../projects/www.worldcommunitygrid.org/wcgrid_opng_autodockgpu_7.28_x86_64-pc-linux-gnu__opencl_nvidia_102 -jobs OPNG_0005115_00189.job -input OPNG_0005115_00189.zip -seed 371057094 -wcgruns 1250 -wcgdpf 25
INFO: Using gpu device from app init data 0
INFO:[04:49:51] Start AutoGrid...

autogrid4: Successful Completion.
INFO:[04:49:55] End AutoGrid...
INFO:[04:49:55] Start AutoDock for ZINC000662752383-ACR2.18_RX1--fr2325benz_003--CYS142.dpf(Job #0)...
OpenCL device: GeForce GT 720
INFO:[04:55:53] End AutoDock...
INFO:[04:55:53] Start AutoDock for ZINC000178506468-ACR2.24_RX1--fr2325benz_003--CYS142.dpf(Job #1)...
OpenCL device: GeForce GT 720
INFO:[04:59:55] End AutoDock...
INFO:[04:59:55] Start AutoDock for ZINC000417995471-ACR2.24_RX1--fr2325benz_003--CYS142.dpf(Job #2)...
OpenCL device: GeForce GT 720
INFO:[05:11:03] End AutoDock...
INFO:[05:11:03] Start AutoDock for ZINC000877352901-ACR2.1_RX1--fr2325benz_003--CYS142.dpf(Job #3)...
OpenCL device: GeForce GT 720
INFO:[05:15:34] End AutoDock...
INFO:[05:15:34] Start AutoDock for ZINC000638390307-ACR2.17_RX1--fr2325benz_003--CYS142.dpf(Job #4)...
OpenCL device: GeForce GT 720
INFO:[05:23:03] End AutoDock...
INFO:[05:23:03] Start AutoDock for ZINC000415459530-ACR2.5_RX1--fr2325benz_003--CYS142.dpf(Job #5)...
OpenCL device: GeForce GT 720
INFO:[05:26:54] End AutoDock...
INFO:[05:26:54] Start AutoDock for ZINC000877356873-ACR2.1_RX1--fr2325benz_003--CYS142.dpf(Job #6)...
OpenCL device: GeForce GT 720
INFO:[05:32:05] End AutoDock...
INFO:[05:32:06] Start AutoDock for ZINC000667207203-ACR2.26_RX1--fr2325benz_003--CYS142.dpf(Job #7)...
OpenCL device: GeForce GT 720
INFO:[05:35:05] End AutoDock...
INFO:[05:35:05] Start AutoDock for ZINC000418171682-ACR2.25_RX1--fr2325benz_003--CYS142.dpf(Job #8)...
OpenCL device: GeForce GT 720
INFO:[05:43:29] End AutoDock...
INFO:[05:43:29] Start AutoDock for ZINC000625810780-ACR2.17_RX1--fr2325benz_003--CYS142.dpf(Job #9)...
OpenCL device: GeForce GT 720
INFO:[05:48:10] End AutoDock...
INFO:[05:48:10] Start AutoDock for ZINC000659632805-ACR2.18_RX1--fr2325benz_003--CYS142.dpf(Job #10)...
OpenCL device: GeForce GT 720
INFO:[05:52:53] End AutoDock...
INFO:[05:52:54] Start AutoDock for ZINC000418224268-ACR2.22_RX1--fr2325benz_003--CYS142.dpf(Job #11)...
OpenCL device: GeForce GT 720
INFO:[05:56:04] End AutoDock...
INFO:[05:56:04] Start AutoDock for ZINC000418186295-ACR2.26_RX1--fr2325benz_003--CYS142.dpf(Job #12)...
OpenCL device: GeForce GT 720
INFO:[23:14:38] End AutoDock...
INFO:[23:14:38] Start AutoDock for ZINC000415431942-ACR2.6_RX1--fr2325benz_003--CYS142.dpf(Job #13)...
OpenCL device: GeForce GT 720
INFO:[23:26:03] End AutoDock...
INFO:[23:26:04] Start AutoDock for ZINC000418005035-ACR2.27_RX1--fr2325benz_003--CYS142.dpf(Job #14)...
OpenCL device: GeForce GT 720
INFO:[23:31:54] End AutoDock...
INFO:[23:31:54] Start AutoDock for ZINC000877356779-ACR2.17_RX1--fr2325benz_003--CYS142.dpf(Job #15)...
OpenCL device: GeForce GT 720
INFO:[23:36:31] End AutoDock...
INFO:[23:36:31] Start AutoDock for ZINC000417972476-ACR2.1_RX1--fr2325benz_003--CYS142.dpf(Job #16)...
OpenCL device: GeForce GT 720
INFO:[23:42:41] End AutoDock...
INFO:[23:42:41] Start AutoDock for ZINC000877257237-ACR2.7_RX1--fr2325benz_003--CYS142.dpf(Job #17)...
OpenCL device: GeForce GT 720
INFO:[23:47:45] End AutoDock...
INFO:[23:47:45] Start AutoDock for ZINC000877453338-ACR2.21_RX1--fr2325benz_003--CYS142.dpf(Job #18)...
OpenCL device: GeForce GT 720
INFO:[23:52:13] End AutoDock...
INFO:[23:52:13] Start AutoDock for ZINC000669533484-ACR2.6_RX1--fr2325benz_003--CYS142.dpf(Job #19)...
OpenCL device: GeForce GT 720
INFO:[23:57:30] End AutoDock...
INFO:[23:57:30] Start AutoDock for ZINC000419869401-ACR2.13_RX1--fr2325benz_003--CYS142.dpf(Job #20)...
OpenCL device: GeForce GT 720

</stderr_txt>
]]>

[Apr 27, 2021 6:20:42 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Ian-n-Steve C.
Senior Cruncher
United States
Joined: May 15, 2020
Post Count: 180
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: I am having a problem with my GPU not always using its full power and the usage is not constant



this is what my GPU usage is doing I have an Nvidia RTX 2080 super and a ryzen 5 1400 CPU a pny 240 GB SSD that can consistently go 500 MB/s I don't know what is wrong and I would like to use my GPU's full power my usage is staying at about 0-5% until suddenly it spikes to 100% for about a second and then back to 0%



Completely normal behaviour. Each spike is basically the GPU working on 1 job. Each GPU task you receive likely has at a couple dozen jobs.

If you want to increase the GPU usage, you could create an app_config file.

not exactly "completely" normal. the old tasks did have this behavior from jobs starting and stopping within the WU, but these new ones, the 5-digit jobs have MUCH more time with the GPU at 0-low%. starkly different between the old and new, and proportionally more run time as a result.
----------------------------------------

EPYC 7V12 / [5] RTX A4000
EPYC 7B12 / [5] RTX 3080Ti + [2] RTX 2080Ti
EPYC 7B12 / [6] RTX 3070Ti + [2] RTX 3060
[2] EPYC 7642 / [2] RTX 2080Ti
[Apr 27, 2021 7:09:34 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Dfirebug
Cruncher
Joined: Jan 6, 2021
Post Count: 5
Status: Offline
Reply to this Post  Reply with Quote 
Re: I am having a problem with my GPU not always using its full power and the usage is not constant

how would I go about making this app configuration file I am good with computers but not great with coding
[Apr 27, 2021 7:53:09 PM]   Link   Report threatening or abusive post: please login first  Go to top 
motech
Cruncher
Joined: Mar 30, 2007
Post Count: 23
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: GPU Work Units - Post Your Tech Support Questions Here

I am receiving multiple GPU WUs for OpenPandemics, however every single one ends with a "Computation error"...
GPU: Nvidia GeForce GT 740, driver version 466.11

You're probably experiencing the same issue that I and others have been having with the GT 7xx series ever since the beta testing. I think it's been established that these cards are able to complete WUs properly on Linux but not on Windows. Although it's entirely possible I may have missed it, I have yet to see a plan to deal with this issue.
----------------------------------------
[Edit 1 times, last edit by motech at Apr 28, 2021 1:49:39 AM]
[Apr 28, 2021 1:45:21 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Michael and Tona
Cruncher
United States
Joined: Oct 13, 2007
Post Count: 4
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: GPU Work Units - Post Your Tech Support Questions Here

Greetings! Did something change from yesterday and today with the Open Pandemics WU's? Yesterday I was crunching them at 2 to 2 1/2 minutes. Today they are 10 to 13 minutes and my GPU is only running roughly 50% of the time on the WU. When the WU hits about 30% the GPU kicks in. I have 2 1080 ti hybrids.

Thanks,

Michael
[Apr 28, 2021 10:52:42 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Eugene Zenzen
Veteran Cruncher
USA
Joined: Mar 31, 2006
Post Count: 888
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: GPU Work Units - Post Your Tech Support Questions Here

how would I go about making this app configuration file I am good with computers but not great with coding
I have this question too.
----------------------------------------

[Apr 29, 2021 7:23:10 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Bigeagle
Cruncher
Joined: Nov 29, 2008
Post Count: 5
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: GPU Work Units - Post Your Tech Support Questions Here

Maybe someone can shed some light on the behavior of the OpenPandemics Covid19 gpu workunits.

I'm running BOINC on Win7, nvidia 1070 and an i5 3570K. I'm monitoring hardware activity and status with hwinfo64 and the msi afterburner.
The gpu is mostly just idle on those WUs and it looks like there is nearly no possible benefit of running the task on the gpu if only a few seconds gpu time come on top of about 10 minutes cpu time. While one could argue that those typical spikes in gpu load is easily misleading in pure visual representation the temperature is a pretty good indirect indicator for power usage which is guaranteed to come with computation. No power usage, no computation. While the typical GPU temperature under load is at about 67-70° C (with games or primegrid WUs) with OpenPandemics Covid19 it was at 53 and since there was quite low activity i would suspect that most of the power consumption came from the gpu just being in high performance state.

So is this an error? Usual behavior of those WUs? Maybe my setup? Or is the GPU part in those just some spice on top of a cpu workload?
It certainly gives me the feeling that this gpu time is wasted. In addition to that those WUs produce distinctly lag on the windows desktop. Which makes the observed idling gpu even more vexing.
----------------------------------------
[Edit 1 times, last edit by Bigeagle at Apr 29, 2021 7:41:27 PM]
[Apr 29, 2021 7:39:19 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Ian-n-Steve C.
Senior Cruncher
United States
Joined: May 15, 2020
Post Count: 180
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: GPU Work Units - Post Your Tech Support Questions Here

Maybe someone can shed some light on the behavior of the OpenPandemics Covid19 gpu workunits.

I'm running BOINC on Win7, nvidia 1070 and an i5 3570K. I'm monitoring hardware activity and status with hwinfo64 and the msi afterburner.
The gpu is mostly just idle on those WUs and it looks like there is nearly no possible benefit of running the task on the gpu if only a few seconds gpu time come on top of about 10 minutes cpu time. While one could argue that those typical spikes in gpu load is easily misleading in pure visual representation the temperature is a pretty good indirect indicator for power usage which is guaranteed to come with computation. No power usage, no computation. While the typical GPU temperature under load is at about 67-70° C (with games or primegrid WUs) with OpenPandemics Covid19 it was at 53 and since there was quite low activity i would suspect that most of the power consumption came from the gpu just being in high performance state.

So is this an error? Usual behavior of those WUs? Maybe my setup? Or is the GPU part in those just some spice on top of a cpu workload?
It certainly gives me the feeling that this gpu time is wasted. In addition to that those WUs produce distinctly lag on the windows desktop. Which makes the observed idling gpu even more vexing.


the app is just poorly optimized now. lots of GPU idle time. you're correct that GPU time is being wasted.

you can help this along by running more GPU tasks concurrently, but at the expense of using more CPU resources which you may or may not want to tolerate. some users are reporting running 2, 4, 8, 10+ concurrent tasks. you need to reserve a CPU core for each additional core. your i5 only have 4 threads. so it'll be pegged 100% supporting 4 concurrent GPU tasks. you won't be able to process anything else on the system.
----------------------------------------

EPYC 7V12 / [5] RTX A4000
EPYC 7B12 / [5] RTX 3080Ti + [2] RTX 2080Ti
EPYC 7B12 / [6] RTX 3070Ti + [2] RTX 3060
[2] EPYC 7642 / [2] RTX 2080Ti
----------------------------------------
[Edit 2 times, last edit by Ian-n-Steve C. at Apr 29, 2021 8:12:36 PM]
[Apr 29, 2021 8:11:17 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 290   Pages: 29   [ Previous Page | 6 7 8 9 10 11 12 13 14 15 | Next Page ]
[ Jump to Last Post ]
Post new Thread