A GPU Tree Database for Many-Core Explicit State Space Exploration

Abstract Various techniques have been proposed to accelerate explicit-state model checking with GPUs, but none address the compact storage of states, or if they do, at cost losing completeness procedure. We investigate how implement a tree database store states as binary trees in GPU memory. present fine-grained parallel algorithms find and trees, experiment number GPU-specific configurations, propose novel hashing technique, called Cleary-Cuckoo hashing, which enables use Cleary compression on GPUs. are first assess effectiveness using database, compression, Experiments show processing speeds up 131 million per second.