Compiler and Runtime Supports for Efficient and Scalable Big Data Systems

Big Data analytics applications such as social network analysis and web analysis have revolutionized modern computing. The processing demand posed by an unprecedented amount of data challenges both industrial practitioners and academia researchers to design and implement highly efficient and scalable system infrastructures. Unfortunately, Big Data processing is fundamentally limited by memory inefficiencies inherent with the underlying programming languages. While offering several invaluable benefits, a managed runtime comes with time and space overheads. In large-scale systems, the memory management cost can be easily magnified and become the critical performance bottleneck. Throughout my Ph.D., I have designed and developed a series of system optimizations to enable scalable Big Data processing including a new programming model, and several novel compiler and runtime supports. In the future, I plan to continue addressing the low-performance issue in data-intensive systems by developing practical solutions across the whole software stack, spanning from the processing model to language extensions.