Jiang Hu - Machine Learning Techniques for Microprocessor Power Modeling and Performance Diagnosis

Abstract: Microprocessor designs require increasingly large manpower and long turn-around time due to the continuous growth of design complexity. This talk introduces two machine learning techniques for fast evaluation of microprocessor power and performance, the two most important design metrics. The first is a power-modelling technique that can reduce power emulation time from weeks to minutes and has been validated in modern industrial designs. The second technique can automate microprocessor performance bug detection, a challenging problem due to the lack of golden reference and has long been a very time-consuming manual process.

Bio: Jiang Hu is a professor in the Department of Electrical and Computer Engineering at Texas A&M University. His research interests include design automation of VLSI circuits and systems, computer architecture optimization and hardware security. He has published over 240 technical papers. He received best paper awards at ACM/IEEE Design Automation Conference 2001, IEEE/ACM International Conference on Computer-Aided Design 2011, IEEE International Conference on Vehicular Electronics and Safety 2018 and IEEE/ACM International Symposium on Microarchitecture 2021. He was the technical program chair and general chair of the ACM International Symposium on Physical Design in 2011 and 2012. He was named an IEEE fellow in 2016.

Видео Jiang Hu - Machine Learning Techniques for Microprocessor Power Modeling and Performance Diagnosis канала Digital Futures: Research Hub for Digitalization