Sebastian Kloibhofer,
"Run-time Data Analysis to Drive Compiler Optimizations"
: SPLASH Companion 2021: Companion Proceedings of the 2021 ACM SIGPLAN International Conference on Systems, Programming, Languages, and Applications: Software for Humanity, ACM Digital, 10-2021
Original Titel:
Run-time Data Analysis to Drive Compiler Optimizations
Sprache des Titels:
Englisch
Original Buchtitel:
SPLASH Companion 2021: Companion Proceedings of the 2021 ACM SIGPLAN International Conference on Systems, Programming, Languages, and Applications: Software for Humanity
Original Kurzfassung:
Dynamic compilers collect a variety of information to optimize programs and achieve peak performance. Nevertheless, particularly in data-heavy applications, analysis of the processed data - data structures, metrics, relations - could enable additional optimizations in terms of access patterns and data locality. Query planning in database systems is one source of inspiration, but due to the required overhead to collect such information, it is infeasible in dynamic compilers. With this project, we propose integrating data analysis into a dynamic runtime to speed up big data applications. The goal is to use the detailed run-time information for speculative compiler optimizations based on the shape and complexion of the data to improve performance.