NSF has awarded Prof. Xipeng Shen a CCF grant on "Exploring and Exploiting Heterogeneous Cache Sharing in CMP Systems for Locality Optimization and Proactive Cache Management."
This 3-year project is focused on cache sharing in modern chip multiprocessors. The sharing of memory hierarchy has become deeper, heterogeneous and more complex, causing cache contention, increased conflicts, and also, synergy sharing. Prior studies, mostly in areas of architecture and operating systems, rely on simple heuristics to estimate cache requirement of corunning programs; the inaccuracy and overhead limits their scalability and effectiveness. This work tackles these challenges uniquely from the compiler aspect by constructing predictive behavior models for corunning processes, developing cache-sharing-aware program transformations and loop scheduling, and combining the program-level knowledge of programming systems with the proactive resource management by runtime systems.