Abstract: Heterogeneous computing has revolutionized the performance of many computing applications. For instance, by harnessing the power of CPU-GPU heterogeneous parallelism, neural network training can be 100 times faster than with CPUs alone. However, developing programs that can effectively use heterogeneous computing resources can be challenging due to the many technical details involved, such as managing parallelism abstraction, accelerator programming, and scheduling. In this talk, I will introduce a new heterogeneous programming system to simplify the creation of high-performance computing applications. I will discuss the key innovations that make our system unique, including a novel programming model, a unified heterogeneous programming solution, and a learning-based runtime. These innovations offer a combination of programmer productivity and performance portability. Finally, I will demonstrate how we have applied our system to accelerate both classical and quantum computing problems. Our system has already been downloaded thousands of times and is being used by various academic and industrial projects.

Bio: Dr. Huang is an assistant professor in the ECE Department at the University of Utah. He received his Ph.D. from the ECE Department at the University of Illinois at Urbana-Champaign and his BS/MS from the CS Department at Taiwan’s National Cheng Kung University. His research group has been creating software systems to simplify the building of high-performance computing applications, including machine learning, computer-aided design, and quantum computing. Dr. Huang has received several awards for his research contributions, including ACM SIGDA Outstanding Ph.D. Dissertation Award, NSF CAREER Award, and Humboldt Research Fellowship Award.

RSVP for Zoom meeting information: https://pitt.co1.qualtrics.com/jfe/form/SV_3Cdob9SUcYdjuS2