Technology

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 knew it! Windows 3.1 number one!!

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.

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.

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.