Seminar: Accuracy-aware Compilers for Energy-efficient Machine Learning
With the slowdown of Moore's law and the end of Dennard scaling, the gap between computer hardware performance and the computational demands of software applications is widening. To close this gap, we need to consider innovative optimizations at all layers of the computing stack. Towards this pursuit, I work on accuracy-aware compilers with the goal to improve the performance and energy-efficiency of software programs. The compilers I build significantly improve the performance and energy usage of programs by systematically trading off small amounts of computational accuracy, often without impacting end-to-end application-level quality.
In this talk, I will discuss ApproxHPVM and ApproxCaliper. ApproxHPVM is an end-to-end compiler framework for accuracy-aware optimization of deep learning workloads running on heterogeneous edge computer platforms. ApproxHPVM is end-to-end in that it only requires high-level information from developers and automatically chooses low-level approximation techniques and knobs that tune the accuracy and performance tradeoffs. The ApproxHPVM compiler supports two different automated approximation selection techniques each with its own merits: i)ApproxTuner (PPoPP'21), and ii) ApproxHPVM IR optimizer (OOPSLA'19).
I will also discuss ApproxCaliper, the first application-aware neural network optimization framework that provides significantly higher speedups compared to application-agnostic tuning methods. I will also lay out a future roadmap of my research vision that focuses on enabling accuracy-aware compilers for a wide range of emerging application domains.
Hashim is a postdoctoral researcher in computer science at the University of Illinois at Urbana-Champaign,working with Vikram Adve and Sasa Misailovic. In April 2021, he completed his PhD from the CS department at the University of Illinois at Urbana-Champaign, advised by Vikram Adve. Hashim works on compilers, systems for machine learning, and program analysis. His PhD thesis focuses on compiler and runtime systems or end-to-end accuracy-aware optimization of tensor-based programs running on heterogeneous edge computer hardware. Hashim's earlier PhD work includes code size reduction techniques via software specialization.
In his research, he has closely collaborated with Industry and Labs,including IBM Research, SRI International, Earthsense, Amazon, and Argonne National Lab. Teams at IBMResearch use the ApproxHPVM compiler infrastructure for compiling programs for a custom heterogeneous SoC. Earthsense (a startup for agriculture robots) uses ApproxCaliper, an application-aware neural network optimization framework for improving the system performance of their resource-constrained agriculture robots.
Hashim is a recipient of the Sohaib and Sara Abbasi Fellowship at the University of Illinois.