Self-Organizing Maps Computing on Graphic Process Unit
Zhongwen Luo, Hongzhi Liu, Zhengping Yang, Xincai Wu · The European Symposium on Artificial Neural Networks · 2005
Self-Organizing Maps (SOM) is a widely used artificial neural network (ANN) model. Because of its heavy computation load when the map is big and inherent parallel, there is a need to apply a parallel algorithm on it. As a SIMD parallel processor, Graphic processing unit (GPU) shows a fast growing speed than CPU. And it also provides programmability recently. In this paper, the algorithm and result of SOM computing on GPU has been given. The result shows that GPU can make SOM computing much faster than standard CPU. Some design tricks for improving the efficiency of computing has discussed. Based on the results and current trends in the development of GPU, it is reasonable to expect that graphic hardware will widely used in other ANN computing for getting high-performance.