<architecture> (Or "working set model") A technique whereby
the operating system in a paging virtual memory
multitasking environment loads all pages of a process's
working set into memory before the process is restarted.
Under demand paging a process accesses its working set by
page faults every time it is restarted. Under prepaging the
system remembers the pages in each process's working set and
loads them into physical memory before restarting the process.
Prepaging reduces the page fault rate of reloaded processes
and hence generally improves CPU efficiency.
["Modern Operating Systems", Andrew S. Tanenbaum,
pub. Prentice Hall, Inc. 1992].
