PENINGKATAN KECEPATAN PEMROSESAN ALGORITMA GENETIKA DENGAN PARALELISASI MENGGUNAKAN METODE COARSE-GRAINED; GENETIC ALGORITHM PROCESSING SPEEDUP WITH PARALELISM USING COARSE-GRAINED METHOD

Rathomi, M Radzi, Reza M.I. Pulungan

2015 | Disertasi | FMIPA

Genetic algorithms are frequently used to solve optimization problems. As the growth of technology, the problem to be solved by genetic algorithms become more complex and require a long time to solve a problem. One solution to speed up the genetic algorithm processing is to use parallelization. Parallelization can be done by dividing the population into sub-populations. Parallelization method that will used is coarse-grained. Parallelization is built with two levels, the first level is executed by message passing interface (MPI) and the second level is executed by using the Graphics Processing Unit (GPU). The test results show that the accuracy of the genetic algorithm which is designed the same as the sequential genetic algorithm. Parallelization with coarsegrained method can improve processing speed of genetic algorithm. Convergence speed of parallel genetic algorithm is better than the sequential genetic algorithm.

Kata Kunci : Genetic Algorithms; Paralelization; Coarse-grained; MPI; GPU.