Samsung Forum

Samsung regularly hosts global leaders and experts to share insights about their cutting edge research and unique experiences. Recordings of these sessions, which happen either at our San Jose campus or now virtually, are shared below. Watch these recordings and learn about trends and cutting edge technology that stand to make a big impact on our future.

October 27, 2020

Design and Design Automation for Efficient ML Hardware Specialization

The growing complexity of machine learning (ML) models (esp. deep neural networks), the proliferation of edge devices, and the diminished benefits of technology scaling together place strong demands on continued improvements in performance and energy efficiency of computer hardware. In line with this trend, ML processing is shifting from general-purpose CPUs/GPUs to specialized architectures in both academic and commercial settings. Increasing specialization also motivates the deployment of new high-level design languages and design automation tools to implement hardware accelerators in a much more productive and agile manner. This talk introduces our recent research on ML hardware specialization, where we investigate both new hardware-friendly ML algorithms and design automation for ML hardware accelerators.

Specifically, we will focus on two topics: (1) Dynamic channel and precision gating, a fine-grained and trainable technique for DNN acceleration. Unlike static network pruning, our approach exploits input-dependent dynamic sparsity at run time. This results in a significant reduction in compute cost with a minimal impact on accuracy. (2) HeteroCL, a new open-source programming framework for accelerator-rich computing. HeteroCL decouples the algorithm from hardware customization techniques to allow much greater flexibility and efficiency in mapping the ML algorithm to custom accelerators.

Zhiru Zhang

Cornell University

Zhiru Zhang is an Associate Professor in the School of ECE at Cornell University. His current research investigates new algorithms, design methodologies, and automation tools for heterogeneous computing. His research has been recognized with a Facebook Research Award (2020), Google Faculty Research Award (2018), the DAC Under-40 Innovators Award (2018), the Rising Professional Achievement Award from the UCLA Henry Samueli School of Engineering and Applied Science (2018), a DARPA Young Faculty Award (2015), and the IEEE CEDA Ernest S. Kuh Early Career Award (2015), an NSF CAREER Award (2015), the Ross Freeman Award for Technical Innovation from Xilinx (2012), and multiple best paper awards and nominations. Prior to joining Cornell, he was a co-founder of AutoESL, a high-level synthesis start-up later acquired by Xilinx.