Definition:

  • Each process gets a small unit of CPU time, i.e., time quantum q, until finish
  • if q is too large, it is similar to first come first serve
  • if q is too small, overhead for context switching
  • No starvation

two