I would say they would need to toss all the 11 scores and do it again without the drive being encrypted (as they acknowledged in the article). Then run modern apps to compare, because ram usage doesn’t say much about speed. Newer games in specific do benchmark better on the same hardware from 10 to 11, showing that while the ram usage is b Higher, it is better used for the time period it is being produced for.

Write speeds and such are up as well apparently.

https://www.techspot.com/article/2349-windows-11-performance/

I also like Buffalo Springfield.

I knew it! Windows 3.1 number one!!

Total RAM usage is a poor benchmark. Windows (and all other major OSes) cache frequently accessed files in free memory to avoid having to read them from disk every time they’re accessed. The cache is freed when the memory is needed, so it doesn’t impact how much memory applications can use, and thus shouldn’t be included in usage benchmarks.

However, Task Manager counts cache memory towards the total usage, which tends to confuse people who think Windows is using way more memory than it actually is. I’m sure the article’s conclusions about memory usage are still mostly correct, but it’s good to keep in mind.

I would assume that newer versions use more RAM, too, because they gradually added features or tweaked configurations that use more RAM. But I don’t see how this changes the reality of the actual usage number…?

You can argue that they’re better utilizing modern hardware and providing features users want, so the usage number should be judged differently. That is definitely the case. But Windows is still the slowest consumer operating system, so it certainly doesn’t absolve them of the critique.

Except that windows 11 isn’t offering the user anything that windows ten didn’t. That’s a big tradeoff for getting absolutely nothing in return but a slower system.

measuring the ram usage on a blank desktop isn’t really a “benchmark” and doesn’t say anything about the OS itself

Benchmark: noun

1. a standard or point of reference against which things may be compared.

And frankly, whatever memory the OS hogs is less memory for applications to hog.

My laptop is 14 years old with a lightweight modern Linux distro. The “OS” - kernel, desktop environment , and system tray apps, a few widgets - uses 800MB of ram when it’s parked at the desktop after startup. Which means the other 15.2GB is available to my applications, and it makes my wheezy old laptop perfectly functional for most things.

I’m burning 10 of 32GB basically idling. Windows caches ALL the things. If I need those extra gigs, I could still use them.

Heavily depends on the edition. I don’t care about 11 so idk if it happens there, but 10 Home has much higher ram usage than LTSC. The lowest I could get on an official iso was 1.6 on a fresh 2019 (i think) ltsc install