Unified Accelerators for Post-Moore Machine Learning

Investment in machine learning infrastructure has exploded due to its success in improving prediction and decision making for a wide range of applications including healthcare, science, agriculture, social media, and entertainment. Similarly, improvements in algorithms have led to unprecedented growth in Deep Neural Network (DNN) model complexity and size, resulting in a commensurate increase in demand for computing. The slowdown in Moore’s Law has pushed both accelerators and high-end GPUs towards narrow number formats to improve logic density, which introduce new challenges for accurate DNN training and inference. Our research aims to bring novel solutions to the challenges introduced by ubiquitous ever-growing DNN models and datasets. Our proposal targets building DNN platforms that are optimal in performance/Watt across a broad class of workloads and improve utility by unifying the infrastructure for both training models and inference tasks.