Andrea Rosà

Andrea Rosà is a researcher at the University of Lugano (USI), Switzerland, with a focus on performance optimization, concurrency, and profiling in the Java Virtual Machine (JVM) ecosystem. His work spans tool development, benchmarking, and parallel computing, often targeting metrics related to task granularity and synchronization. He has contributed to major conferences such as SPLASH, APLAS, and GPCE, and is actively involved in program committees and artifact evaluations.

His research includes designing profiling tools like P3 and Renaissance, which analyze parallel applications, lock-free operations, and dynamic program behavior. These tools have been instrumental in benchmarking modern JVM performance and identifying variability in workloads.

Andrea has authored numerous papers on profiling Java Streams, vector APIs, and multi-language dynamic analysis. His work often addresses challenges in automated benchmarking, runtime transitions, and task granularity optimization.