Get ready for a truly revolutionary talk on bridging the software-hardware gap! Professor Michael O’Boyle of the University of Edinburgh will be sharing his insights on the exciting new compiler tools and approaches at the Bayes Coffee House Future Informatics Talks seminars.

Title: New compiler tools for a heterogeneous world: Bridging the software hardware gap

Speaker: Michael O’Boyle

                   University of Edinburgh

Time: 10/02(Wed) 11:00-12:00 (UTC+01:00)London

Where: Bayes Centre G03

External: https://app.huawei.com/wmeeting/join/94537256/hSlwHIL0GdU4SMx5cf77kCLdWM6GoLz0u

Meeting ID: 94537256

Passcode: 159654

Abstract: With the decline in transistor scaling, computer systems are increasingly heterogeneous, specialised and diverse. As it stands, software will simply not fit and current compiler technology is struggling to bridge the gap. This talk argues that we need to fundamentally rethink the design of the compiler and consider new approaches from other fields.

This talk presents a range of novel approaches to bridging the software-hardware gap. These include using program synthesis to lift legacy programs to MLIR, an emerging infrastructure for building high-level compilers, that can effectively target modern hardware. An alternative approach uses neural machine translation to compiler and decompile code. It outperforms state of the art decompilers and shows significant improvement over chatGPT.

Bio: Michael O’Boyle is a professor of computer science at the University of Edinburgh. He is best known for his work on incorporating machine learning into compilation and parallelization. He has published over 150 papers, receiving seven best paper and three test of time awards. Over his career he has been awarded three fellowships, five international visiting positions and 30+ grants. He is a fellow of the BCS.

【BAYES COFFEE HOUSE Future Informatics Talks】New compiler tools for a heterogeneous world: Bridging the software hardware gap / Huawei-Edinburgh Joint Lab by blogadmin is licensed under a Creative Commons Attribution CC BY 3.0