cancel
Showing results for 
Search instead for 
Did you mean: 
etherjack
New member
3 1 0 0
Message 1 of 1
167
Flag Post

Interrupt Storm.

HP Recommended

I have a ZBook 17 G5 less than a year old with an identical issue. There was a BIOS update release just a couple weeks ago. I applied that and it seemed to help...sort of. I can now reproduce AND mitigate the issue at will, but the problem doesn't completely go away. I am running a fully patched W10 Enterprise system.

To reproduce:
I run any app that puts a sustained load on my GPU. This could be a graphics-intensive game or some length video rendering process. After several minutes of this activity, I begin to see the utilization of CPU 0 (and ONLY CPU 0) steadily increase. As soon as the GPU activity stops, CPU 0 utilization increases to nearly 100% and stays there indefinitely. The #1 consumer of CPU in this state are system interrupts.

To mitigate (not fix):
I put the PC to sleep, wait a few seconds, and wake it back up. The core in question drops to a normal level and does not have the problem again...until I leverage my GPU again.

Additional discoveries:
One reproducible oddity I found was related to the NVIDIA control panel. I am not a gamer but recently noticed that, when playing a game on my laptop for the first time, the GPU performance was abysmal (6-10 FPS). Through a bit of trial and error I discovered that any effort to change the Performance Counter permissions (even if I don't actually change them) will trigger a "Do you want to save your changes?" dialog. When clicking 'Yes' to that dialog, all my screens go dark for a few moments and then come back. After going through that process, the game easily holds 60FPS with my GPU at only 60%.

 
 

The Performance Counter settings in the NVIDIA control panelThe Performance Counter settings in the NVIDIA control panel

However...

This causes the aforementioned issue with CPU 0 so, when I done playing, I have to put my system through a sleep/wake cycle to correct that issue. If I fail to do so, the CPU temp eventually hits tjmax and system shuts down.
 
In conclusion, my testing suggests this issue is directly related to the GPU configuration and, perhaps, the GPU security settings. This is further supported by a recent firmware update that specifically stated it was designed to address a potential security issue with the video card.

 

Warning Be alert for scammers posting fake support phone numbers and/or email addresses on the community. If you think you have received a fake HP Support message, please report it to us by clicking on "Flag Post".
† The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation