Abstract
Recent work introduced the cube-and-conquer technique to
solve hard SAT instances. It partitions the search space into cubes using a
lookahead solver. Each cube is tackled by a conflict-driven clause learning
(CDCL) solver. Crucial for strong performance is the cutoff heuristic
that decides when to switch from lookahead to CDCL. Yet, this offline
heuristic is far from ideal. In this paper, we present a novel hybrid solver
that applies the cube and conquer steps simultaneously. A lookahead
and a CDCL solver work together on each cube, while communication
is restricted to synchronization. Our concurrent cube-and-conquer solver
can solve many instances faster than pure lookahead, pure CDCL and
offline cube-and-conquer, and can abort early in favor of a pure CDCL
search if an instance is not suitable for cube-and-conquer techniques.
Original language | English |
---|---|
Title of host publication | Proc. Intl. Workshop on Pragmatics of SAT (POS'12), |
Number of pages | 14 |
Publication status | Published - 2012 |
Fields of science
- 102011 Formal languages
- 102 Computer Sciences
- 101 Mathematics
JKU Focus areas
- Computation in Informatics and Mathematics