High Performance Distributed Deep Learning: A Beginner's Guide

A tutorial to be presented at PPoPP 2018: Principles and Practice of Parallel Programming 2018

When: February 25, 2018
Where: Vienna, Austria

Dhabaleswar K. (DK) Panda, Ammar Ahmad Awan, and Hari Subramoni

The Ohio State University


The current wave of advances in Deep Learning (DL) has led to many exciting challenges and opportunities for Computer Science and Artificial Intelligence researchers alike. Modern DL frameworks like Caffe/Caffe2, TensorFlow, CNTK, Torch, and several others have emerged that offer ease of use and flexibility to describe, train, and deploy various types of Deep Neural Networks (DNN) including deep convolutional nets. In this tutorial, we will provide an overview of interesting trends in DL and how cutting-edge hardware architectures are playing a key role in moving the field forward. We will also present an overview of DL frameworks from an architectural as well as a performance standpoint. Most DL frameworks have utilized a single GPU to accelerate the performance of DNN training and inference. However, approaches to parallelizing the process of training are also being actively explored. The DL community has moved along MPI based parallel/distributed training as well. Thus, we will highlight new challenges for MPI runtimes to efficiently support DNN training. We highlight how we have designed efficient communication primitives in MVAPICH2 to support scalable DNN training. Finally, we will discuss how co-design of the OSU-Caffe framework and MVAPICH2 runtime enables scale-out of DNN training to 160 GPUs.

Tutorial Objectives

Recent advancements in Artificial Intelligence (AI) have been fueled by the resurgence of Deep Neural Networks (DNNs) and various Deep Learning (DL) frameworks like Caffe, Facebook Caffe2, Google TensorFlow, and Microsoft Cognitive Toolkit (CNTK). DNNs have found widespread applications in classical areas like Image Recognition, Speech Processing, Textual Analysis, as well as areas like Cancer Detection, Medical Imaging, and even Autonomous Vehicle systems. Two driving elements can be attributed to the momentum that DL has gained recently; first is the public availability of various data sets like ImageNet, CIFAR, etc., and second is the widespread adoption of data-parallel hardware like GPUs and accelerators to perform DNN training. The raw number crunching capabilities of GPUs have significantly improve DNN training. Today, the community is designing better, bigger, and deeper networks for improving the accuracy through models like AlexNet, GoogLeNet, Inception v3, and VGG. The models differ in the architecture (number and type of layers) but share the common requirement of faster computation and communication capabilities of the underlying systems. Based on these trends, this tutorial is proposed with the following objectives:

Targeted Audience

This tutorial is targeted for various categories of people working in the areas of Deep Learning and MPI-based distributed DNN training on modern HPC clusters with high-performance interconnects. Specific audience this tutorial is aimed at include:

Audience Prerequisites

There is no fixed pre-requisite. As long as the attendee has a general knowledge in HPC and Networking, he/she will be able to understand and appreciate it. The tutorial is designed in such a way that an attendee gets exposed to the topics in a smooth and progressive manner. The content level will be as follows: 60% beginner, 30% intermediate, and 10% advanced.

Detailed Outline of the Tutorial

The tutorial is organized along the following topics with a detailed time budget (half-day):

Brief Bio of Presenters

Dr. Dhabaleswar K. (DK) Panda is a Professor and University Distinguished Scholar of Computer Science at the Ohio State University. He obtained his Ph.D. in computer engineering from the University of Southern California. His research interests include parallel computer architecture, high performance computing, communication protocols, files systems, network-based computing, and Quality of Service. He has published over 400 papers in major journals and international conferences related to these research areas. Dr. Panda and his research group members have been doing extensive research on modern networking technologies including InfiniBand, HSE and RDMA over Converged Enhanced Ethernet (RoCE). His research group is currently collaborating with National Laboratories and leading InfiniBand and 10GigE/iWARP companies on designing various subsystems of next generation high-end systems. The MVAPICH2 (High Performance MPI over InfiniBand, iWARP and RoCE) open-source software package, developed by his research group, are currently being used by more than 2,825 organizations worldwide (in 85 countries). This software has enabled several InfiniBand clusters (including the 1st one) to get into the latest TOP500 ranking. More than 430,000 downloads of these libraries have taken place from the project's site. These software packages are also available with the stacks for network vendors (InfiniBand and iWARP), server vendors and Linux distributors. The RDMA-enabled Apache Hadoop, Spark and Memcached packages, consisting of acceleration for HDFS, MapReduce, RPC, Spark and Memcached, are publicly available from High-Performance Big Data (HiBD) project site: http://hibd.cse.ohio-state.edu. These packages are currently being used by more than 250 organizations in 31 countries. More than 23,450 downloads have taken place from the project's site. Dr. Panda's research is supported by funding from US National Science Foundation, US Department of Energy, and several industry including Intel, Cisco, SUN, Mellanox, QLogic, NVIDIA and NetApp. He is an IEEE Fellow and a member of ACM. More details about Dr. Panda, including a comprehensive CV and publications are available here.

Ammar Ahmad Awan Ammar Ahmad Awan received his B.S. and M.S.degrees in Computer Science and Engineering from National University of Science and Technology (NUST), Pakistan and Kyung Hee University (KHU), South Korea, respectively. Currently, Ammar is working towards his Ph.D. degree in Computer Science and Engineering at The Ohio State University. His current research focus lies at the intersection of High Performance Computing (HPC) libraries and Deep Learning (DL) frameworks. He previously worked on a Java-based Message Passing Interface (MPI) and nested parallelism with OpenMP and MPI for scientific applications. He has published 14 papers in conferences and journals related to these research areas. He actively contributes to various projects like MVAPICH2-GDR (High Performance MPI for GPU clusters, OMB (OSU Micro Benchmarks), and HiDL (High Performance Deep Learning). He is the lead author of the OSU-Caffe framework (part of HiDL project) that allows efficient distributed training of Deep Neural Networks.

Dr. Hari Subramoni is a research scientist in the Department of Computer Science and Engineering at the Ohio State University, USA, since September 2015. His current research interests include high performance interconnects and protocols, parallel computer architecture, network-based computing, exascale computing, network topology aware computing, QoS, power-aware LAN-WAN communication, fault tolerance, virtualization, big data and cloud computing. He has published over 50 papers in international journals and conferences related to these research areas. He has been actively involved in various professional activities in academic journals and conferences. Dr. Subramoni is doing research on the design and development of MVAPICH2 (High Performance MPI over InfiniBand, iWARP and RoCE) and MVAPICH2-X (Hybrid MPI and PGAS (OpenSHMEM, UPC and CAF)) software packages. He is a member of IEEE. More details about Dr. Subramoni are available here.

Last Updated: October 26, 2017