Skip to Main Content
Research in task scheduling algorithm is a challenging problem for high-performance computing; especially achieving a better makespan is a key issue in design and development of heterogeneous algorithm. In this paper, we investigate an existing scheduling algorithm and consider the problem of minimizing its makespan. We have formulated our problem as a nonlinear programming model, and then used the genetic algorithm to solve that problem. Experimental results show that the makespan of our new algorithm is proved shorter than those of the two existing scheduling algorithms, while still maintaining the algorithm's high availability.