A framework for compositional transformations of recursions and loops
Frameworks like the polyhedral model provide sound, compositional reasoning for transformations of loop programs. But there is no analogous framework that lets programmers reason about compositions of transformations on recursive programs, despite substantial recent work on specific transformations for these programs. In this talk, I will discuss recent work in my group on techniques for reasoning about transformations of programs that feature nested loops and recursion. I will present new representations of such programs and their transformations, a framework for composing these transformations, and a new, decidable dependence test for proving that the composed transformations are sound.
Milind Kulkarni is an associate professor in the School of Electrical and Computer Engineering at Purdue University, where he is a University Faculty Scholar. His research focuses on developing languages, compilers and systems that can efficiently and effectively exploit locality and parallelism in complex applications on complex computation platforms. He received his Ph.D. in Computer Science from Cornell University in 2008, where he was a Department of Energy High Performance Computer Science (HPCS) Fellow. He received an NSF CAREER award in 2012, a Department of Energy Early Career Research Award in 2013, and a Presidential Early Career Award for Scientists and Engineers in 2016 for his work on optimizing irregular applications. He is the recipient of numerous departmental, college, and university-wide teaching awards.