Gist: efficient data encoding for deep neural network training

TLDR: This paper investigates widely used DNNs and finds that the major contributors to memory footprint are intermediate layer outputs (feature maps), and introduces a framework for DNN-layer-specific optimizations that significantly reduce this source of main memory pressure on GPUs.

Abstract: Modern deep neural networks (DNNs) training typically relies on GPUs to train complex hundred-layer deep networks A significant problem facing both researchers and industry practitioners is that, as the networks get deeper, the available GPU main memory becomes a primary bottleneck, limiting the size of networks it can train In this paper, we investigate widely used DNNs and find that the major contributors to memory footprint are intermediate layer outputs (feature maps) We then introduce a framework for DNN-layer-specific optimizations (eg, convolution, ReLU, pool) that significantly reduce this source of main memory pressure on GPUs We find that a feature map typically has two uses that are spread far apart temporally Our key approach is to store an encoded representation of feature maps for this temporal gap and decode this data for use in the backward pass; the full-fidelity feature maps are used in the forward pass and relinquished immediately Based on this approach, we present Gist, our system that employs two classes of layer-specific encoding schemes -- lossless and lossy -- to exploit existing value redundancy in DNN training to significantly reduce the memory consumption of targeted feature maps For example, one insight is by taking advantage of the computational nature of back propagation from pool to ReLU layer, we can store the intermediate feature map using just 1 bit instead of 32 bits per value We deploy these mechanisms in a state-of-the-art DNN framework (CNTK) and observe that Gist reduces the memory footprint to upto 2X across 5 state-of-the-art image classification DNNs, with an average of 18X with only 4% performance overhead We also show that further software (eg, CuDNN) and hardware (eg, dynamic allocation) optimizations can result in even larger footprint reduction (upto 41X)