The aim of this study was to integrate the simplicity of structured sparsity into existing vector execution flow and vector processing units (VPUs), thus expediting the corresponding matrix ...