In the (current) sample solution, if the time quantum is the same as the time when a system call occurs, the system call doesn't get executed first, and the process is taken off the CPU and the next system call time is set to 0usecs. In the workshop you said that it should be the other way around. I vaguely remember that when the sample solution first came out, the system call was executed first. I was wondering if the sample solution is undergoing continuous changes which is messing things up somehow. I also do agree with the other post about there being an extra usec when there shouldn't be.