Windows 10 furor????

[petrw1]

Ever seen this before? The green leaf...
Other than increasing the priority how do I tell it to stop suspending it?

During this period the Hard drive was running overtime ; though there were no obvious big processes running.

Based on the Prime95 window it suspended either as soon as it started Stage 2 on a new assignment...or when I moved the mouse to turn on the monitor.

Prime95 resumed about 15 minutes later.

[retina]

The performance of the system must be preserved at all times.

How else can MS guarantee to keep that spy data flowing freely to their servers?

Oh, you thought it was your system they were talking about, hahaha, no it is their exploitation system that they are maintaining the performance of.

[rogue]

Do you have anti-virus running?

[pepi37]

A green leaf icon will appear next to that process in Task Manager to indicate it's running in Efficiency mode. Note that Efficiency mode isn't available if you select a core Windows process or a process group. And that you can remove Efficiency mode from any process that is currently using it.

[rogue]

There are also efficiency cores on newer CPUs. It might be defaulting to one of those.

[Chuck]

Originally Posted by rogue

There are also efficiency cores on newer CPUs. It might be defaulting to one of those.

I had a related issue with a new machine and Windows 11. The system kept hibernating after an hour even though in my power plan it was set to sleep "never". I finally stopped this behavior by setting time to sleep "never" in all power plans including the one I was using. I also ran "powercfg.exe /hibernate off" in an elevated cmd window.

To stop the GPU from downclocking, in the NVIDIA control panel I selected "prefer maximum performance" in manage 3D settings. (Don't know if an AMD GPU would be subject to downclocking).

I have had no issues since.

[petrw1]

Happened again today.
First picture is during slowdown.
Second is after reboot when it was running normally again.

Something going crazy with the disk.

[kruoli]

How much main memory was occupied when you took that screenshot? It might be thrashing.

[petrw1]

Originally Posted by kruoli

How much main memory was occupied when you took that screenshot? It might be thrashing.

It was after I stopped/started Prime95 and it was starting stage 2.
It should have taken a couple minutes but this was about 15 minutes later.
I had 13.5/16GB allocated for S2 and I watched it in task manager but instead of climbing quickly to 13.5 it climbed to about 10GB then bounced up and down a little for the whole time between 10 and 11.
Thr whole time CPU usage never got above 6%.

[kriesel]

Originally Posted by petrw1

it was starting stage 2.
It should have taken a couple minutes but this was about 15 minutes later.
I had 13.5/16GB allocated for S2 and I watched it in task manager but instead of climbing quickly to 13.5 it climbed to about 10GB then bounced up and down a little for the whole time between 10 and 11.
Thr whole time CPU usage never got above 6%.

I have found on a 24GB system that allowing 18GB in s2 is too much; 17 is stable, for ordinary wavefront exponents, but still gets slow/paging with 100Mdigit & above. Allowing all but 2.5GB as you appear to be is I think MUCH too aggressive. Try 10 of 16. Or 9.5. And/or be ruthless about exiting everything else. CPU will necessarily be rather idle while waiting for paging out/in of some of the prime95 stage 2 data that does not all fit in available ram.

Also, check the system event log for any indication of disk read or other disk errors. The disk data rates don't look very high. If there is an issue with reliability causing retry delays and error logging, you'll want to know about that in time to do a full backup before the disk becomes completely unusable.

[petrw1]

Originally Posted by kriesel

I have found on a 24GB system that allowing 18GB in s2 is too much; 17 is stable, for ordinary wavefront exponents, but still gets slow/paging with 100Mdigit & above. Allowing all but 2.5GB as you appear to be is I think MUCH too aggressive. Try 10 of 16. Or 9.5. And/or be ruthless about exiting everything else. CPU will necessarily be rather idle while waiting for paging out/in of some of the prime95 stage 2 data that does not all fit in available ram.

Also, check the system event log for any indication of disk read or other disk errors. The disk data rates don't look very high. If there is an issue with reliability causing retry delays and error logging, you'll want to know about that in time to do a full backup before the disk becomes completely unusable.

This computer (and a few others) are Prime95 exclusively.
Nothing else runs on them; okay I have AVG for Virus and occasionally open Google Chrome.
I run them all within 2GB-2.5GB of full and have never had issues with any others.
My family workhorse PC allocates only 24/32GB.

Disk errors seems possible; though in 4 months this PC in question has "paused" 4 times.
As noted in an earlier post the first 2 times I was away and it just "continued" on its own a few days later.
The last 2 times I rebooted and it worked fine for another month.
Stopping and starting Prime95 did not help; I had to reboot.