How pages loaded in RAM other than after page-fault trap?

Question

I am trying to understand demand paging concepts and can't find an answer. How page-fault rate can be relatively small, if only way is for OS to understand that it need particular page, is to fetch instruction and find out that this page is not loaded into RAM? If page-fault rate is small, how OS manages to load in RAM right pages? How it guesses in advance, that particular page will be needed? Thanks.

Answer 1

A demand paging system needs to fetch each initialized page from disk at least once. Once the page is loaded, it only needs to be unloaded if there is a demand for physical memory.

In the days of multi-user systems with 4GB address spaces and 8MB of physical memory, there was a lot of paging.

Now that 8GB of memory is pretty much standard on a single user system, there is not as much paging as there once was.