Loom: exploiting weight and activation precisions to accelerate convolutional neural networks.

Sayeh Sharify, Alberto Delmas Lascorz, Kevin Siu, Patrick Judd, Andreas Moshovos

Browse the full DAC paper archive.