Two Optimizations on the Stålmarck Procedure

Abstract

In this paper, we introduce StalmarckSAT, the a modern re-implementation of the Stålmarck Procedure for SAT solving, and present two novel strategies to improve the Procedure, Cardinality Driven Branching (CDB) and Deductive Priority Ordering (DPO). CDB is a heuristic to improve branching with the dilemma rule, and DPO intelligently orders simple rules based on their deductive potential. Our results demonstrate improved solve times with both strategies.

Presented at the FMCAD 2025 Student Forum. Not part of the official FMCAD proceedings