My research focuses on code compilation challenges, such as optimization for both power efficiency and performance. I address these challenges through codesign of compilers and the hardware platforms they target. My research on compilers has enhanced both compile-time and run-time efficiency.
My current work uses static and dynamic compilation, run-time optimization, and advanced code analysis to extract coarse-grained parallelism for many-core architectures from general purpose sequential code. I have started the HELIX research project in 2010 to demonstrate the potential of this research.
I received my Ph.D. degree with the highest honors from Politecnico di Milano in 2009. I am the author of ILDJIT, an open source parallel compilation framework demonstrating principles from my dissertation. ILDJIT includes both static and dynamic compilers. It is used on several research projects to investigate new approaches to program introspection, optimization, and micro-architectural design.
- IEEE Micro 2012: HELIX: Making the Extraction of Thread-Level Parallelism Mainstream
- HELIX research project
Past and current collaborators: