GPU usage at ~3% on RTX2070
3 Oct 2021 11:58:44 UTC
Topic 226132
(moderation:
Aloha,
I found some threads with this topic but none seems to have such low usage. It's never over 12%. What did I do wrong? Temperature also says there's not much done with my GPU :´(
Currently: gravitational wave research, GPU active. CPU (AMD 3950x 16 core) at 100%.
Must note, too: The progress for the GPU part is mostly stuck at 99%? (!)
greetings
Markus
P.S. I downloaded newest
)
P.S. I downloaded newest NVIDIA studio drivers
Patience. At the end of the
)
Patience. At the end of the Workunit, there is some post-processing going on that is only done by the CPU.
To better keep your GPU usage high, you can always run several tasks at the same time. Staggered starts and multiple tasks at the same time is the best to get this
Aside from the benefits of
)
Aside from the benefits of running multiple GPU tasks simultaneously, you may find benefit by greatly reducing competing usage of your CPU (your GPU needs frequent support from the CPU throughout the task, and if it has to wait for it--utilization goes down and elapsed time goes up).
An easy way to check whether competing BOINC CPU tasks are slowing you down is just to suspend all available CPU tasks, then monitor activity through a couple of task cycles. If you get a big benefit, you can try lowering the allowable number CPU tasks, or adjusting task priorities, or other tricks of the trade.
Also, you'll probably like your results better if you don't mix the two flavors of GPU tasks (Gravity Wave and Gamma-Ray Pulsar).
Since your expressed concern is with the GW GPU tasks, I suggest as a trial you suspend ALL other BOINC tasks save only for GW GPU, let them run a couple of cycles, and see if it makes a difference through the full task cycle.
As the GPU utilization varies greatly in the course of a single task, useful comparisons are best made using GPU utilization averaging. Either GPU-Z or HWiNFO supports that.
Thanks guys for the detailed
)
Thanks guys for the detailed ansers!
Changed preferences to 2 GPU tasks and my GPU starts working ;)
How can I do staggered starts for GPU? All tasks are synced right now. Thanks in advance
Markus Windisch wrote:How can
)
If you have tasks in queue already: Suspend a task somewhere in the middle while it's running. Then another task should start. It's not easy to give any suggestions when exactly would be the best moment to suspend a running task. Just try pausing them... and you will find good intervals.
Or you could set "no new tasks", let your queue run out, open up computing preferences and set 0 for "store at least X days of fork" and "store up to and additional X days of work". Then set "allow new tasks" again. Right after the first task begun running... set "no new tasks". Then wait... and at some point set "allow new tasks" again. Boinc should download another one and start it.
If you have tasks in queue
)
I must say it's also working
)
I must say it's also working with 4 tasks. One of these tasks is always a little faster now so I don't have to do anything. Will experiment a little and post my results just in case someone finds this thread.
Markus Windisch wrote:I
)
I found it to be too much trouble to constantly manage the queues after each reboot. The gamma ray pulsar workunits are way easier to manage (just works) and also reward more credits.
Many Nvidia gpus do not seem
)
Many Nvidia gpus do not seem to run more than one gamma ray task at a time effectively.
This might be true of gravity wave tasks too.
The Nvidia GPU manager for Windows will tell you how loaded your GPU is.
So will gpu-z.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote: Many Nvidia
)
So will Windows 10 Task Manager. Select Performance tab and select there Nvidia GPU. There you see 4 different categories of loads. Click on the title of one of them (arrow down) and select Cuda. This will show the computing load (includes OpenCL as well) of that GPU.