Nearly all big science, machine learning, neural network, and machine vision applications employ algorithms that involve large matrix-matrix multiplication. But multiplying large matrices pushes the ...