Implementing a Reduction Clause to Overcome Critical Section Deficiencies in Parallel Computing

Parallel programming is a method we can use to increase processor performance in parallel computers because almost all multicore computers integrate more than one individual processor and cache memory. There are several techniques that can be used in parallel computing, one of which is by utilizing the Critical Section to protect the occurrence of data inconsistencies in the parallel computing process. There are advantages and disadvantages to using Critical Section, namely Mutually Exclusive where no work can be done simultaneously in a Critical Section on one parallel program. So in this journal we will try to discuss solutions to protect data inconsistencies in parallel computing without having to experience Mutually Exclusive, namely by using the Reduction Clause method. The results of this study show that Reduction Clause can function without the need for a waiting process and offset parallel looping with correct results without data inconsistency and also increase parallel computing work time when compared to us using Critical Section.