The Sinclair QL Forum

Use Task Manager instead of Resource Monitor, go to CPU tab and active "Show kernel times" under "View" I guess.

And wow, apparently I had hair once

mk79 wrote:Use Task Manager instead of Resource Monitor, go to CPU tab and active "Show kernel times" under "View" I guess.

I switched back and forth a couple of times. At the point of the arrow, I started a screen saver that always blew one core to close to 100%. The rest is just idling. The fan kicks in after about 3 mins of idling, which it never did before..

And wow, apparently I had hair once

Hair is vastly overrated

pjw wrote:I switched back and forth a couple of times. At the point of the arrow, I started a screen saver that always blew one core to close to 100%. The rest is just idling. The fan kicks in after about 3 mins of idling, which it never did before..

Hm ok, no idea. Obviously QPC didn't change in the last year, so something else must have. I did expect much more kernel time (the red line), which usually hints at a hardware problem. Does this happen without anything loaded in SMSQ/E as well?

Marcel

Definitely time to open up that laptop. Also, is it mining cryptocurrency without your informed consent?

mk79 wrote:

pjw wrote:I switched back and forth a couple of times. At the point of the arrow, I started a screen saver that always blew one core to close to 100%. The rest is just idling. The fan kicks in after about 3 mins of idling, which it never did before..

Hm ok, no idea. Obviously QPC didn't change in the last year, so something else must have. I did expect much more kernel time (the red line), which usually hints at a hardware problem. Does this happen without anything loaded in SMSQ/E as well?

Ahem. I should have tested more thoroughly before rending my cloak and pouring ashes on my head in public Without anything loaded the fan doesnt come on after three minutes (or ten for that matter) The CPU strain is still nearly 100% on one of the cores - in fact with the tools I have available I cant tell the difference between a loaded and an empty system. So my fully loaded system must just be pushing it over the edge. My conclusion therefore is that either Im running something new thats burning CPU cycles, or that the machine needs a thorough clean.. Ill do an inventory asap. The cleaning will have to wait for some boring, wet afternoon.
Thanks for your help! And sorry for bothering everyone with this false alarm. (I had a number of good reasons for my worry, but should have done some basic testing before asking for assistance.)

Dave wrote:Definitely time to open up that laptop. Also, is it mining cryptocurrency without your informed consent?

Thats hard to tell, isnt it? I was more concerned with Spectre and Meltdown, as I recently found out that they were discovered about the time my machine started playing up, although the info only hit the public sphere recently. Its not the exploits that are the problem so much as the cure which, Im told, could significantly affect performance on older CPUs. This may still be the problem here (I know virtually nothing about these things, and thats how I like to keep it). But I think your suggestion to open up and give her a good blowout is a good one, and thats what I intend to do.

On my Mac the patch has made, at most, a 2-3% difference.

I would just watch task by task in Activity Monitor and look at the "energy impact" column. There are some crypto loads that will attempt to mine, and try to disguise their loads as some random task. There's also some weird peaks that come from housekeeping for the new AppleFS filing system, especially if it is encrypted.

But yes I'd be opening the machine up and eliminating a physical cause too, since that's just something you do every now and then anyway and it is due, based on your previous statements.

pjw wrote:Ahem. I should have tested more thoroughly before rending my cloak and pouring ashes on my head in public Without anything loaded the fan doesnt come on after three minutes (or ten for that matter) The CPU strain is still nearly 100% on one of the cores - in fact with the tools I have available I cant tell the difference between a loaded and an empty system.

That is still strange, QPC doesn't consume even 1% of CPU when idling here on my machine. You could download Process Explorer, right click QPC and hit "Properties". Then go to the "threads" tab and see which thread consumes the QPC.

Marcel

mk79 wrote:

pjw wrote:Ahem. I should have tested more thoroughly before rending my cloak and pouring ashes on my head in public Without anything loaded the fan doesnt come on after three minutes (or ten for that matter) The CPU strain is still nearly 100% on one of the cores - in fact with the tools I have available I cant tell the difference between a loaded and an empty system.

That is still strange, QPC doesn't consume even 1% of CPU when idling here on my machine. You could download Process Explorer, right click QPC and hit "Properties". Then go to the "threads" tab and see which thread consumes the QPC.

To be frank, I no longer remember what was "normal" behaviour; I only started reacting to the noise. I do remember that my "Matrix" screen saver maxed out the one core, leading to this sort of reacket - even when QPC2 was running in the background. Which is why I abandoned it as a screen saver and just sometimes use it as an aid to meditation.. (Only joking).
Anyway, Process Explorer wont play, as it complains over a missing M$ Symbol Server. (It appears M$'s customers are also moaning over the missing symbols, to which M$'s most recent response (20.03.2018) is "The current delays we’re hitting are due to the last piece of our legacy pipeline that we’re on track to replace in the next month." Any recommended alternatives would be appreciated..

Whoops! It suddenly decided to play along, after all. Here goes:
Drat! I posted from the wrong machine! The next post will (hopefully) show the correct one.