QEA를 이용한 멀티코어에의 작업 할당

Along with the advances in semiconductor process technologies, multi-core architectures are getting more popular. As the number of cores integrated on a chip increases, task mapping is becoming one of the most important problems to be solved for an effective use of those cores. In this paper, we adopt the quantum-inspired evolutionary algorithm to solve the mapping problem and propose an efficient encoding scheme that exploits the characteristics of the problem. Experiments show that the scheme achieves 40% power reduction compared to a naive encoding scheme.