• Corpus ID: 199442467

Learned Clause Minimization in Parallel SAT Solvers

  title={Learned Clause Minimization in Parallel SAT Solvers},
  author={Marc Hartung and Florian Schintke},
Learned clauses minimization (LCM) let to performance improvements of modern SAT solvers especially in solving hard SAT instances. Despite the success of LCM approaches in sequential solvers, they are not widely incorporated in parallel SAT solvers. In this paper we explore the potential of LCM for parallel SAT solvers by defining multiple LCM approaches based on clause vivification, comparing their runtime in different SAT solvers and discussing reasons for performance gains and losses… 

Figures and Tables from this paper


An Effective Learnt Clause Minimization Approach for CDCL SAT Solvers
A new inprocessing SAT approach is defined which eliminates redundant literals from learnt clauses by applying Boolean constraint propagation and shows that a remarkable number of additional instances are solved when the approach is incorporated into five of the best performing CDCL SAT solvers.
Tuning Parallel SAT Solvers
New implementation details and benchmarking results for the authors' parallel portfolio solver \topo are presented, and the impact of diversifying the solver by using different restart- , branching- and clause database management heuristics are discussed.
A parallel portfolio SAT solver with lockless physical clause sharing
A parallel SAT solver that allows for sharing clauses between several threads logically and physically and shows how physical clause sharing can be used to propagate one thread’s improvements on the clause database to all solving threads.
Lazy Clause Exchange Policy for Parallel SAT Solvers
A simple parallel version of Glucose is proposed that uses a lazy policy to exchange clauses between cores that does not send a clause when it is learnt, but later, when it has a chance to be useful locally.
Partitioning SAT Instances for Distributed Solving
New methods for constructing partitions which combine clause learning and lookahead are presented and their performance is demonstrated with an extensive comparison against the best sequential solvers in the SAT competition 2009 as well as against two efficient parallel solvers.
Concurrent Clause Strengthening
This work extends a conventional CDCL SAT solver with a second computation thread, which is solely used to strengthen the clauses learned by the solver, which provides a simple and natural way to exploit the availability of multi-core hardware.
Between SAT and UNSAT: The Fundamental Difference in CDCL SAT
This paper identifies direct connections to the workings of some of the most important elements in CDCL solvers: the effects of restarts and VSIDS, and the roles of learned clauses, and gives a wide range of concrete evidence that highlights the varying effects and roles of these elements.
An Extensible SAT-solver
This article presents a small, complete, and efficient SAT-solver in the style of conflict-driven learning, as exemplified by Chaff, and includes among other things a mechanism for adding arbitrary boolean constraints.
Refining Restarts Strategies for SAT and UNSAT
This paper explores restart strategies in the light of a new vision of SAT solvers, and considers CDCL solvers as resolution-based producers of clauses, and shows how detecting sudden increases in the number of variable assignments can help the solver to target SAT instances too.
Cube-and-Conquer for Satisfiability
This chapter presents the cube-and-conquer paradigm, a competitive alternative for solving SAT problems in parallel which outperforms both lookahead and conflict-driven solvers and achieves linear-time speedups on this problem even when using thousands of cores.