A high-performance predictor for critical unstable generators (CUGs) of power systems is presented in this paper. The predictor is driven by the MapReduce based parallelized neural networks. Specifically. a group of back propagation neural networks (BPNNs). https://www.adaptsurvival.com/product-category/pasta/