A novel accelerator named DyCNN is proposed for sparse convolutional neural network(CNN)that has the in-creasing scale and rapid evolution.DyCNN is an energy-efficient and flexible accelerator,which is based on coarse grained reconfigurable architecture(CGRA).DyCNN utilizes a data-aware dynamic filtering mechanism to elimi-natea large number of invalid calculations and memory accesses caused by the static sparsity of filters and dynamic sparsity of activation values in sparse convolutional neural network and increase the on-chip reuse of instructions among processing units.Meanwhile,a dynamic work-stealing strategy combined with a static work distribution scheme is proposed to alleviate the load imbalance caused by the sparsity of filter and activation values.Overall,DyCNN achieves a 1.69 × speedup and 3.04 × energy savings on average when running sparse CNN compared with running dense CNN.DyCNN achieves 2.78 ×,1.48 × speedup and 35.62 × and 1.17 × energy savings com-pared with the state-of-the-art GPU(cuSPARSE)and Cambricon-X solutions respectively.
维普
万方数据