Improving Computer Processing Efficiency by Sequencing

Suppose that a set of n programming jobs is given to a computer for processing, where the processing times are independent random variables. It is known that sequencing jobs based on shortest expected service time is optimal. In this paper, we study the efficiency improvement of sequencing computer jobs for processing in an order as predicted by an expert, such as a programming wizard. We get closed form analytical results for the expectation and variance of the total system time under random versus ordered sequencing and we compare the results. We analyze the amount by which the expert’s ordering is better than the random order rule.