Hey guys!

Been doing this for a short bit now, and was wondering if there was a way to increase GPU load. I run GPUgrid and Seti@home (as well as WCG and another) and was wondering if there was a way to increase my GPU load. My GPU is only running at about 2-10% utilization at any time and would love to try to get more umph out of my GTX 1080 Ti when I'm not gaming.

Anyone know how to do so?

Thanks for any help!

Sincerely,
Kiyometa

There is a setting that is especially potent on Linux called SWAN_SYNC. This parameter also helps on Windows but not to quite the extent as on Linux. I don't have time at the moment to explain how to enable it but perhaps someone else can.

My GPU is only running at about 2-10% utilization at any time

I guess you have this readout from Windows task manager's GPU tab. You should change the view of one of the graphs to "CUDA" to get the right values. Judging by the task's runtimes on your host you should have about 90% GPU utilization for "long run" tasks. "Short run" tasks may yield lower GPU utilization, as these are intended for low end cards. Preferably you should run "long run" tasks on your GTX 1080Ti. (You can change it in your preferences at the GPUGrid webpage, even though this part is broken.)

... and would love to try to get more umph out of my GTX 1080 Ti when I'm not gaming.
Anyone know how to do so?

You can try to reduce the number of CPU tasks running (if you run any of them) preferably to 8 (or 7, if you use SWAN_SYNC), or even less.

You can try to apply the swan_sync environmental value:
Press Windows key + R
Type sysdm.cpl and press [enter]
Click on the "Advanced" tab (the middle one)
Click on [Environmental Variables] (it's near the lower right corner)
Look for the lower section called "System Variables", click on the [New] button below the list of "System Variables".
Type swan_sync in the name field
Type 1 in the Value field
Click [OK] 3 times.
Exit BOINC manager with stopping scientific applications.
Start BOINC manager.

You should also create an app_info.xml file for BOINC manager to notice that the GPUGrid app is using 1 full CPU.
The app_config.xml file should be placed to the project's home directory (by default it's at c:\ProgramData\BOINC\projects\www.gpugrid.net\)
Copy the following to the clipboard:

notepad c:\ProgramData\BOINC\projects\www.gpugrid.net\app_config.xml

Press Windows key + R, then paste and press enter.
Copy & paste the following text:

<app_config>
 <app>
  <name>acemdlong</name>
  <gpu_versions>
   <gpu_usage>1</gpu_usage>
   <cpu_usage>1</cpu_usage>
  </gpu_versions>
 </app>
<app>
  <name>acemdshort</name>
  <gpu_versions>
   <gpu_usage>1</gpu_usage>
   <cpu_usage>1</cpu_usage>
  </gpu_versions>
 </app>
</app_config>

Click file -> save and click [save].
Exit BOINC manager with stopping scientific applications.
Start BOINC manager.
(If your BOINC manager is running, you can click Options -> read config files.)

Sweet, got it peaking to 100% now.
Thanks for the help!

Hello,

I try your solution but it does not change anything.
I also add one more node in the xml app file:

<app_config>
 <app>
  <name>acemdlong</name>
  <gpu_versions>
   <gpu_usage>1</gpu_usage>
   <cpu_usage>1</cpu_usage>
  </gpu_versions>
 </app>
<app>
  <name>acemdshort</name>
  <gpu_versions>
   <gpu_usage>1</gpu_usage>
   <cpu_usage>1</cpu_usage>
  </gpu_versions>
 </app>
 <app>
  <name>acemd3</name>
  <gpu_versions>
   <gpu_usage>1</gpu_usage>
   <cpu_usage>1</cpu_usage>
  </gpu_versions>
 </app>
</app_config>

Because I have this issue messsages:
https://i.ibb.co/6Dn4xZZ/GPUGRID-acemd.jpg


Best Regards,

Wilgard

acemd3 is the only project running now. The other two are extinct: acemdshort & acemdlong.
Not sure what you're trying to solve.
BTW, those warning messages may also appear just because you have no WUs of that type downloaded. Just ignore it.

Yes, just delete the app_config file for the project. It is not needed for just the one app that is available.

Ok, but in this case, what should I have to change to "increase GPU utilization" ?

the best way is to reduce the CPU tasks that are running.

also your laptop with the GT620M GPU in it is too old to work on this project.

---

Hi Ian&Steve C,
How do you do that ?
I ask for the other laptop with NVIDIA Quadro M520.

set your cpu % in the compute preferences to below 100%. set it to 80 or something.

---

I use mutiple projects:
WCG
LCH
GPUGRID
Rosetta
CPDN
So when CPU preference is at 100% I have 4 workunits which work like :
- 1 GPUGRID
- 1 CWG
- 2 Roretta
If I have to 80% one of them is set in pause mode, but the GPU part of the WU of gpugrid does not change at all.

you have to make a compromise somewhere. using all of your CPU for other tasks is starving the GPUGRID task of resources. without a full CPU thread feeding the GPU, you will suffer performance loss like you are seeing. you're trying to run 5 jobs that all want a full CPU thread, with only 4 CPU threads.

you will either have to deal with the low GPU utilization, or accept that you will have to stop processing work on another CPU thread.

---

In fact I don't want to do any compromise so having a CPU to 100% and a GPU to 100% too.
But as you said i have to do a compromise so I prefer the current state thus a lower GPU usage.
Thank you