Download PDFOpen PDF in browser

Exploiting parallelism in the ME calculus

13 pagesPublished: August 19, 2013

Abstract

We present some parallelization techniques for the Model Evolution (ME) calculus,
an instantiation-based calculus that lifts the DPLL procedure
to first-order clause logic.

Specifically, we consider a restriction of ME to the EPR fragment
of clause logic for which the calculus is a decision procedure.
The main operations in ME's proof procedures, namely clause instantiation
and candidate literal generation, offer opportunities for MapReduce-style
parallelization.

This term/clause-level parallelization is largely orthogonal to the sort
of search-level parallelization performed by portfolio approaches.
We describe a hybrid parallel proof procedure for the restricted calculus
that exploits parallelism at both levels to synergistic effect.
The calculus and the proof procedure have been implemented in
a new solver for EPR formulas.

Our initial experimental results show that our term/clause-level parallelization
alone is effective in reducing runtime and can be combined
with a portfolio-based approach to maximize performance.

Keyphrases: instantiation based calculi, model evolution, parallel theorem proving

In: Pascal Fontaine, Renate A. Schmidt and Stephan Schulz (editors). PAAR-2012. Third Workshop on Practical Aspects of Automated Reasoning, vol 21, pages 96-108.

BibTeX entry
@inproceedings{PAAR-2012:Exploiting_parallelism_ME_calculus,
  author    = {Tianyi Liang and Cesare Tinelli},
  title     = {Exploiting parallelism in the ME calculus},
  booktitle = {PAAR-2012. Third Workshop on Practical Aspects of Automated Reasoning},
  editor    = {Pascal Fontaine and Renate A. Schmidt and Stephan Schulz},
  series    = {EPiC Series in Computing},
  volume    = {21},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2398-7340},
  url       = {/publications/paper/WdR},
  doi       = {10.29007/v1f2},
  pages     = {96-108},
  year      = {2013}}
Download PDFOpen PDF in browser