Do you know exactly what type of measurement you're looking at when you say Linux is using lots of memory? Accounting for memory usage is actually a lot more complicated than you might think. For example in Task Manager, Windows splits memory usage in to the categories "In use", "In use (compressed)", "Modified", "Standby" and "Free", with further measurements for "committed", "paged pool", and "non-paged pool". The concept of "total memory use" depends on which of these categories you include. For example in Windows the "standby" category is used in the sense there is lots of potentially useful cached data and code stored in memory which can help optimize the use of the system; however if an application suddenly demanded lots more memory, it can generally just release all that memory and give it to the application. So it's not used in the sense it blocks other applications from allocating memory. On my system memory is 41% full if you count just "In use", but 92% full if you count both "In use" and "Standby"; only about 7% is actually categorised as "Free". So the measurements can be wildly different depending on what is being counted, and the meaning of the result is significantly different too: 92% in use might be a problem, but 92% in use and standby is not a problem, because 51% of that memory is still actually available to applications should they need it.