Prof. Yufei Ding

We name our lab to highlight PICASSO's spirit of life-long creation and self-surmounting. We believe "research is like art". On one hand, both researchers and artists need to master a set of skills to make our work attractive and appealing. On the other hand, what really matters, in the end, is what we want to say to the world.

The long-term research goal of the PICASSO group is to build automatic, intelligent programming systems that could leverage, integrate, and advance math, physics, and computer science. The target territory to realize our research dream includes Machine Learning (ML) and Quantum Computing (QC). Our recent efforts cut across multiple programming system technologies, ranging from program verification and testing to high-level algorithmic optimization and autotuning, to domain-specific programming language designs, kernel library implementations, advanced compilation constructions, and computer architecture designs.

In particular, two key research principles lay the foundation for our research: "abstraction" and "automation". The former is the key enabler for many high-level and large-scope optimization opportunities that were not possible within a single technical stack, and it is also our muse when exploring cross-stack synergy and integration. The latter also encodes our research philosophy: researchers shall focus on original work rather than "incremental" work in slightly changed settings. One of the key ingredients for avoiding such repeated work is automation.