EBPC: Extended Bit-Plane Compression for Deep Neural Network Inference and Training Accelerators

TL;DR

This paper introduces a hardware-efficient, lossless compression scheme for CNN feature maps that significantly reduces I/O bandwidth and energy consumption, improving performance for neural network inference and training accelerators.

Contribution

The paper presents a novel, hardware-friendly compression method with hardware architectures, achieving high compression ratios and low silicon area for CNN accelerators.

Findings

Achieves an average compression ratio of 5.1x for AlexNet

Reduces silicon area to less than seven 8-bit multiply-add units

Improves compression ratios for training gradient maps beyond inference

Abstract

In the wake of the success of convolutional neural networks in image classification, object recognition, speech recognition, etc., the demand for deploying these compute-intensive ML models on embedded and mobile systems with tight power and energy constraints at low cost, as well as for boosting throughput in data centers, is growing rapidly. This has sparked a surge of research into specialized hardware accelerators. Their performance is typically limited by I/O bandwidth, power consumption is dominated by I/O transfers to off-chip memory, and on-chip memories occupy a large part of the silicon area. We introduce and evaluate a novel, hardware-friendly, and lossless compression scheme for the feature maps present within convolutional neural networks. We present hardware architectures and synthesis results for the compressor and decompressor in 65nm. With a throughput of one 8-bit word/cycle at 600MHz, they fit into 2.8kGE and 3.0kGE of silicon area, respectively - together the size of less than seven 8-bit multiply-add units at the same throughput. We show that an average compression ratio of 5.1x for AlexNet, 4x for VGG-16, 2.4x for ResNet-34 and 2.2x for MobileNetV2 can be achieved - a gain of 45-70% over existing methods. Our approach also works effectively for various number formats, has a low frame-to-frame variance on the compression ratio, and achieves compression factors for gradient map compression during training that are even better than for inference.