- This event has passed.
Defense Notification: Tobi Popoola
February 24 @ 1:30 pm - 3:30 pm MST
Title: Sparse Format Conversion and Code Synthesis
Program: Doctor of Philosophy in Computing
Advisor: Dr. Catherine Olschanowsky, Computer Science
Committee Members: Dr. Elena Sherman, Computer Science; and Dr. Tim Andersen, Computer Science
Sparse computations are important in scientific computing. Many scientific applications compute on sparse data. These computations are considered irregular because they introduce programming patterns not suited for regular compiler optimization techniques. Sparse formats make use of auxiliary arrays to store non-zeros, as a result, the contents of auxiliary arrays are not known until run-time. The Inspector/Executor (I/E) paradigm enables the use of run-time information for compiler optimizations. An inspector computes information at run-time to drive transformations. The executor-a compile-time transformation of the original code- uses information computed by the inspector. The sparse polyhedral framework (SPF) encompasses a series of tools to support I/E run-time transformations. This work introduces a unified framework that wraps SPF tools while providing a holistic view of a computation as an intermediate representation (IR). This work also introduces a method to automatically synthesize inspectors for transforming between sparse formats and improvements SPF to explore performance of irregular applications.