Windows 10 Interrupts taking high CPU at idle

[KittX]

This problem has been happening eventually for a while, now it's getting really annoying.
If i leave PC for a few moments, the CPU gets 100% load and Process Explorer shows Interrupts taking 75-80% of the CPU usage.
I've been testing with plugging/unplugging all usb devices, extra sata hdds etc; been trying unplugging the rig monitor (i've got 2 - at the rig and at the desk), and couple weeks ago we came to conclusion it might be SSD with OS failing. I've replaced the SSD, migrated my OS to the new one and havent had any problems eversince, for 2 weeks, and today it's back.
We've got a few central heating issues at the moment so its usually quite cold at night, and i've had some driving and racing in the rig yesterday evening, so its got quite warm inside the case. But then i left it idling into the night, and temp in the room dropped for quite a bit. I have checked the system with antiviruses just in case (also these interrupts issues don't happen in safe mode), all is clean.
Are there any diagnostic tools for motherboard to diagnose any short circuits or bad contacts?
Basically once you left pc for a few minutes, you cannot do anything on it for about 10 minutes onwards; until the Interrupts CPU usage goes down again, everything is just in slideshow mode.

PC Specs: Motherboard ASUS Maximus VIII Ranger, I7 6700K (not OC), GTX1080, PSU Seasonic 760W, Fractal Design high tower case with Noctua fans all around, usually stuff doesn't get too warm in there, GPU not exceeded 73C so far yet. But i bet there can be quite a drop in temperature when games are running vs just idling. At idle it's stone cold.

 

[Ho3n3r]

Run Malwarebytes. It sounds like you have some sort of cryptominer or any type of malware that targets your CPU.

Also monitor your clock speeds with CPU-Z or HWMonitor or the like to make sure it doesn't go through the roof temlerature-wise, reaching thermal throttling temperature.

Also, make sure your anti-virus is up to date and running.

 

[KittX]

Thanks, I've ran Malwarebytes, also showed nothing. Oh, and it goes with that high interrupts on idle, if i unplug network cable ase well.

 

[RasmusP]

Not sure if that helps in this case but you could try to use "Latency Mon" to find out what's causing it.
I use it to get behind issues with audio dropouts/spikes when recording music. These dropouts are directly connected to CPU load. If I run latency mon these audio spiking goes through the roof, as it's testing and measuring exactly that. Also the CPU load increases and I just checken: process explorer Interrupts are going from 0.6% cpu load to 3%.

When you stop the measuring you can see the statistics about interrupts. Not sure if it's the same "interrupts" but I for example can see that the nvidia driver, usb driver and my audio card driver are causing the biggest amount of lags.

 

[dud]

/looks at Asus and their BIOS "skills".

Any chance there is a new BIOS available? Or if not are you willing to try an older one?

 

[KittX]

Cheers guys so far for the help. During yesterday's testing we've discovered that if you unplug one of memory sticks from A2 B2 slots it works fine, on any of those sticks. On any of those slots :O.

I've then put memory to A1 B1 and the PC wasnt booting, the LED panel on the motherboard was showing the "memory identification failed" code. That was fixed by unplugging cpu_fan cable and booting. Then plugged cpu_fan cable back into socket, and it boots now. Noticed that XMP was disabled after that, maybe its been disabled when the memory was failed to identify in A1B1 slots. Enabled XMP, memory stays in A1 B1 for now, everything is okay so far... will keep monitoring if it goes back or not. Id like to put it back to A2B2 eventually.

 

[dud]

That really smells like BIOS fubar.

Glad that you found a solution.

 

[KittX]

I still tend to think some unsecure seating/bad contacts or cables can induce some voltage fluctuations so the high interrupts.
For example I noticed I tend to not put memory sticks into sockets hard enough. Low side of stick, where the socket doesnt have clamp, sometimes are too sticking outside compared to the side with clamps.
Also noticed my rig display's displayport cable connection is also somewhat faulty, not sure if it is socket or plu in the cable. Will test another cable asap.
But i have ran tests yesterday without rig monitor eonnected at all, and there were still high interrupts (before my manipulations with memory).
Oh well, lets see how it goes.

 

[dud]

RAM when working does not involve interrupts.

It is hard to imagine how badly seated RAM legitimately causes high interrupt load (except maybe a bad module with ECC, but that isn't the case here).

 

[KittX]

Just bumping the thread, as this gets really annoying. Interrupts were coming back every once in 1-2 weeks since then, and every time reseating the components or just changing the internal configuration or plugging out one of displays worked.
I once thought it's the onboard soundcard giving circuits (as it was giving a big static noise when you plug soundsystem cables to it... I bought pci-e soundcard, disabled internal one and interrupts have gone... for about a month. Now they're back.
Seems like they always are coming back after extensive work/load to the pc, the next time you launch it. Re-configuring internal/external components seems to fix that temporarily.