Solving N-Queens problem on GPU architecture using OpenCL with special reference to synchronization issues

The N-Queens problem is to place N queens on an N × N chessboard such that no two queens attack each other. General purpose computing on graphics processing units (GPGPU) is fast becoming a common feature of high performance computing. This paper investigates cost of finding the solutions to N-Queens problem on GPGPU architecture using OpenCL programming model. We extensively analyze the N-Queen problem with respect to local, global memory parameters and atomicity and synchronization issues in OpenCL and its effects on performance. Experimental results are shown on NVidia Quadro FX 3800 GPU. Using Queens between 16 and 21, we observed average speedup of 20x.