2 ms·Scalable Near-Memory Architecture for Training Deep Neural Networks In-Memory2 points by godelmachine 9y ago