Publication detail

GPU-Accelerated Synthesis of Probabilistic Programs

ANDRIUSHCHENKO, R. ČEŠKA, M. MARCIN, V. VOJNAR, T.

Original Title

GPU-Accelerated Synthesis of Probabilistic Programs

Type

conference paper

Language

English

Original Abstract

We consider automated synthesis methods for finite-state probabilistic programs satisfying a given temporal specification. Our goal is to accelerate the synthesis process using massively parallel graphical processing units (GPUs). The involved analysis of families of candidate programs is the main computational bottleneck of the process. We thus propose a state-level GPU-parallelisation of the model-checking algorithms for Markov chains and Markov decision processes that leverages the related but distinct topology of the candidate programs. For structurally complex families, we achieve a speedup of the analysis over one order of magnitude. This already leads to a considerable acceleration of the overall synthesis process and paves the way for further improvements.

Keywords

Markov models, probabilistic programs, graphical processing units

Authors

ANDRIUSHCHENKO, R.; ČEŠKA, M.; MARCIN, V.; VOJNAR, T.

Released

30. 6. 2022

Publisher

Springer Nature Switzerland AG

Location

Cham

ISBN

978-3-031-25312-6

Book

International Conference on Computer Aided Systems Theory (EUROCAST'22)

Edition

Lecture Notes in Computer Science

Pages from

256

Pages to

266

Pages count

11

BibTex

@inproceedings{BUT178306,
  author="Roman {Andriushchenko} and Milan {Češka} and Vladimír {Marcin} and Tomáš {Vojnar}",
  title="GPU-Accelerated Synthesis of Probabilistic Programs",
  booktitle="International Conference on Computer Aided Systems Theory (EUROCAST'22)",
  year="2022",
  series="Lecture Notes in Computer Science",
  pages="256--266",
  publisher="Springer Nature Switzerland AG",
  address="Cham",
  isbn="978-3-031-25312-6"
}