- Local time
- 3:41 AM
- Posts
- 1,992
- OS
- Windows 10 Pro
What I meant was 'the actual RAM usage of the OS as measured by the Task Manager'. For example, the RAM usage of the OS so measured on the Raspberry PI and how that grows over the years. And the RAM usage of the OS so measured on the laptop and how that grows over the years. Exactly as how Bree did it, i.e. compare the machine to itself over the years. Don't compare different machines (Raspberry PI vs laptop) that is not a valid comparison.You can't measure anything with the Task Manager. You can only see how much is used/given to services and it is usually more than they really need. If you have little RAM the OS use little RAm and if you have a lot of extra RAM the OS will give away extra/bigger chunks of it for smoother operation of these services.
Example: I ran Windows 10 Pro on a Raspberry Pi 3 with only 1GB of RAM. It consumed as liitle as 350MB. While on my Laptop with 32GB of RAM the usage was closer to 6-8GB with exactly the same setup. And nowhere in the Task Manager you can see how much any of the services/processes really were using, only how much were reserved/given to them.
Wow, your data is complete, resulting in only 0.4GB increase of actual RAM usage of the OS over the years on your test machine. As I wrote further up, I think that the small but inevitable bloat contributes only little to the slow down of older computers. The removable 'crud' contributes more to the slow down of older computers, but this contribution is reversible (and there are non-reversible slow downs, too)I'm in a position to look at that, having a test machine with system images of every version of W10, identical software installed in each.
Yes, when the system has been left to settle down, and with no software open, my 21H1 uses more RAM than 1507 (10240), but only by about 400MB.
My Computer
System One
-
- OS
- Windows 10 Pro