Accelerating Sparse Matrix Operations in Neural Networks on Graphics Processing Units

37045

Knowledge Triples