high performance computing on graphics processing units: hgpu.org

The increasing complexity of deep neural networks (DNNs) has made it challenging to exploit existing large-scale data process pipelines for handling massive data and parameters involved in DNN training. Distributed computing platforms and GPGPU-based acceleration provide a mainstream solution to this computational challenge. In this paper, we propose DeepSpark, a distributed and parallel deep learning framework that simultaneously exploits Apache Spark for large-scale distributed data management and Caffe for GPU-based acceleration. DeepSpark directly accepts Caffe input specifications, providing seamless compatibility with existing designs and network structures. To support parallel operations, DeepSpark automatically distributes workloads and parameters to Caffe-running nodes using Spark and iteratively aggregates training results by a novel lock-free asynchronous variant of the popular elastic averaging stochastic gradient descent (SGD) update scheme, effectively complementing the synchronized processing capabilities of Spark. DeepSpark is an on-going project, and the current release is available.