Find Academic Supervisors

Andrea RosàView profile

Researcher

• Dynamic Program Analysis
• Managed Languages and Runtimes
• Dynamic Compilers
• +7 more

Andrea Rosà is a Researcher and Lecturer at the Faculty of Informatics of Università della Svizzera italiana (USI) in Lugano, Switzerland. He is a member of the Dynamic Analysis Group (DAG), where he conducts research on managed language runtimes, dynamic program analysis, and performance optimization of parallel systems. Andrea earned his PhD in Informatics from USI in 2018 under the supervision of Prof. Walter Binder. Prior to that, he completed both his MSc and BSc in Computer Science Engineering at Politecnico di Milano. His research centers on dynamic program analysis, managed language runtimes (especially the Java Virtual Machine), dynamic compilers, benchmarking, concurrency and parallelism, and big-data analytics. He has developed tools such as AkkaProf for profiling actor-based applications and contributed to the Renaissance benchmarking suite. His recent work investigates task granularity, Java streams, vectorization, and compiler-level profiling. His most recent publications (2023–2025) explore JVM performance, Java Vector API, native-image startup, and adaptive threading, published in leading venues such as TOPLAS, PLDI, CGO, and OOPSLA. The articles collectively reflect a strong focus on empirical systems research, performance engineering, and tool development for modern JVM-based applications. Andrea collaborates extensively with researchers including Matteo Basso, Walter Binder, Eduardo Rosales, and Lydia Y. Chen. He has co-authored over 50 publications and contributed to multiple open-source projects and benchmarking frameworks. He advises no known students but actively contributes to academic service, including organizing and editing workshop proceedings. He is based at the East Campus, Sector D, Office D5.10, Via la Santa 1, Lugano-Viganello, Switzerland.

• Parallel Computing
• Java Virtual Machine (JVM) Optimization
• Performance Profiling
• +5 more

Walter Binder is a Full Professor in the Faculty of Informatics at the Università della Svizzera italiana (USI). He holds a MSc, PhD, and venia docendi from Vienna University of Technology. Previously, he was a senior researcher at the Artificial Intelligence Laboratory, EPFL. His research focuses on program transformations, virtual execution environments, aspect-oriented programming, profiling, and resource management. Education: MSc from Vienna University of Technology PhD from Vienna University of Technology Venia docendi (Habilitation) from Vienna University of Technology Research Interests: His work spans JVM optimization, parallel computing, and dynamic program analysis. Notable contributions include tools like Renaissance (a JVM benchmark suite), S2S (SQL-to-Stream translator), and Akkaprof (profiler for actor-based systems). He explores topics such as thread management, vectorization, and performance profiling in distributed systems. Articles Trends: Recent work emphasizes JVM performance (e.g., Java Vector API, Native Image optimizations), parallel execution policies (NAS benchmarks), and adaptive runtime systems (e.g., MPR framework). He also investigates big data systems (Spark, Node.js) and runtime verification tools. Grants & Labs: Leads projects on JVM profiling, polyglot runtimes, and large-scale program analysis. His lab develops tools like NodeMOP for Node.js and AccStream for stream processing systems.