Elena Favero
An AI Agent System to Generate JMH Microbenchmarks.
Rel. Luca Ardito, Tommaso Fulcini, Philipp Leitner. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Informatica (Computer Engineering), 2026
|
Preview |
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (3MB) | Preview |
Abstract
Code performance is currently one of the key aspects of software development, since a performance inefficiency results in longer waiting times and higher costs, aspects that should be reduced as much as possible. Measuring code performance becomes essential and one way to evaluate it are microbenchmarks, performance tests that focus on individual instructions or methods. While several frameworks exist to support developers in writing them, including JMH for Java code, writing correct and meaningful microbenchmarks by hand remains a difficult and time-consuming task. Recent research has therefore attempted to automate this process, although existing approaches present certain limitations, such as a reliance on existing test cases or a lack of tools available to LLMs.
This project aims to provide a tool capable of automating the entire benchmark generation process
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Corso di laurea
Classe di laurea
Ente in cotutela
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
