Efficient and portable GEMM-based convolution operators for deep neural network training on multicore processors

Download PDF
Open Webpage

Sergio Barrachina, Manuel F. Dolz, Pablo San Juan, Enrique S. Quintana-Ortí

Published: 2022, Last Modified: 15 May 2025J. Parallel Distributed Comput. 2022EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•We present DECONVGEMM, an operator integrating the col2im transform and the BLIS GEMM.•We present a reindex transform and a transposed CONVGEMM for gradient computation.•We provide implementation descriptions of the novel CONVGEMM/DECONVGEMM operators.•We integrate our operators in PyDTNN: Python Distributed Training of Neural Networks.•We evaluate the performance and memory savings of the CONVGEMM/DECONVGEMM operators.