GraphChallenge.org Sparse Deep Neural Network Performance

Abstract

The MIT/IEEE/Amazon GraphChallenge.org encourages community approaches to developing new solutions for analyzing graphs and sparse data. Sparse AI analytics present unique scalability difficulties. The Sparse Deep Neural Network (DNN) Challenge draws upon prior challenges from machine learning, high performance computing, and visual analytics to create a challenge that is reflective of emerging sparse AI systems. The sparse DNN challenge is based on a mathematically well-defined DNN inference computation and can be implemented in any programming environment. In 2019 several sparse DNN challenge submissions were received from a wide range of authors and organizations. This paper presents a performance analysis of the best performers of these submissions. These submissions show that their state-of-the-art sparse DNN execution time, $T_{\mathrm{DNN}}$, is a strong function of the number of DNN operations performed, $N_{\mathrm{op}}$. The sparse DNN challenge provides a clear picture of current sparse DNN systems and underscores the need for new innovations to achieve high performance on very large sparse DNNs.

arXiv comment7 pages, 7 figures, 80 references, to be submitted to IEEE HPEC 2020. This work reports new updated results on prior work reported in arXiv:1909.05631. arXiv admin note: substantial text overlap with arXiv:1807.03165, arXiv:1708.02937. arXiv admin note: text overlap with arXiv:2003.09269

References in corpus (8)

• SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
• Community structure and scale-free collections of Erdös-Rényi graphs
• Streaming Graph Challenge: Stochastic Block Partition
• Speeding up Convolutional Neural Networks By Exploiting the Sparsity of Rectifier Units
• Using a Power Law Distribution to describe Big Data
• PageRank Pipeline Benchmark: Proposal for a Holistic System Benchmark for Big-Data Platforms
• Sparse Deep Neural Network Exact Solutions
• On Large-Scale Graph Generation with Validation of Diverse Triangle Statistics at Edges and Vertices

Cited by in corpus (2)

• Activation Sparsity Opportunities for Compressing General Large Language Models
• Training for temporal sparsity in deep neural networks, application in video processing