Academic Journal

USING THE SPIN MODEL CHECKER FOR AUTO-TUNING HIGH-PERFORMANCE PROGRAMS.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: USING THE SPIN MODEL CHECKER FOR AUTO-TUNING HIGH-PERFORMANCE PROGRAMS.
Συγγραφείς: Gorlatch, Sergei1 (AUTHOR) gorlatch@uni-muenster.de, Garanina, Natalia2,3 (AUTHOR), Staroletov, Sergey4 (AUTHOR)
Πηγή: Journal of Mathematical Sciences. May2025, Vol. 290 Issue 5, p641-653. 13p.
Θεματικοί όροι: *High performance computing, *Parallel programs (Computer programs), *OpenCL (Computer program language), *Software verification, *Program transformation, *Mathematical optimization, *Modeling languages (Computer science), *Model validation
Περίληψη: The paper bridges two traditionally separate research directions: (1) model checking, commonly used in the formal verification of programs, and (2) auto-tuning, frequently applied in high-performance computing. Auto-tuning frameworks optimize parallel programs by identifying the optimal values of performance-critical parameters—known as tuning parameters—for specific high-performance architectures and data sizes. Given the large number of parameters affecting a program's performance, determining the optimal configuration is a challenging task, even for experts. Although auto-tuning automates this process, it is often time-intensive. We accelerate auto-tuning by applying model checking, utilizing a counterexample generated during the verification of the program's optimality property. This paper provides a detailed implementation of our approach for OpenCL programs—a standard for programming modern high-performance architectures—using the Promela modeling language and the SPIN verification tool, along with a report on our experimental results. [ABSTRACT FROM AUTHOR]
Βάση Δεδομένων: Academic Search Index
Περιγραφή
ISSN:10723374
DOI:10.1007/s10958-025-07687-3