A Review on GPU Based Parallel Computing for NP Problems

Abstract

Now a days there are different number of optimization problems are present. Which are NP problems to solve this problems parallel metaheuristic algorithm are required. Graph theories are most commonly studied combinational problems. In this paper providing the new move towards solve this combinational problem with GPU based parallel computing using CUDA architecture. Comparing those problem with relevant to the transfer rate, effective memory utilization and speedup etc. to acquire the paramount possible solution. By applying the different algorithms on the optimization problem to catch the efficient memory exploitation, synchronized execution, saving time and increasing speedup of execution. Due to this the speedup factor is enhance and get the best optimal solution.