Acceleration of hierarchical Bayesian network based cortical models on multicore architectures

We examine the parallelization of two biologically inspired hierarchical Bayesian cortical models onto recent multicore architectures. These models have been developed recently based on new insights from neuroscience and have several advantages over traditional neural network models. In particular, they need far fewer network nodes to simulate a large scale cortical model than traditional neural network models, making them computationally more efficient. This is the first study of the parallelization of this class of models onto multicore processors. Our results indicate that the models can take advantage of both thread and data level parallelism to provide significant speedups on multicore architectures. MPI implementations on clusters of multicore processors were also examined, and showed that the models scaled well with the number of machines in the clusters.