Spintronics-Compatible Approach to Solving Maximum Satisfiability Problems with Probabilistic Computing, Invertible Logic and Parallel Tempering

By Andrea Grimaldi et al

Read the original document by opening this link in a new tab.

Table of Contents

Abstract
Introduction
Ising Machines
Probabilistic Computing
Max-SAT
Invertible Logic and Max-SAT Instance
Annealing with Parallel Tempering

Summary

The document discusses an approach to solving maximum satisfiability problems using spintronics, probabilistic computing, invertible logic, and parallel tempering. It introduces a scalable approach combining probabilistic computing with p-bits, parallel tempering, and invertible logic gates for facing Max-SAT problems. The study demonstrates the spintronic implementation of this approach based on Landau-Lifshitz-Gilbert equations. It benchmarks the algorithm with hard Max-SAT instances from the 2016 competition, showing potential energy-efficient and fast architecture for COP solutions. The work explores the use of PC for solving Max-SAT, weighted Max-SAT, and Max-Cut problems, emphasizing the potential of spintronic technology for compact hardware implementation. The document also covers the application of invertible logic gates in mapping combinatorial problems and the use of parallel tempering in the annealing process for minimizing unsatisfied clauses. The study provides insights into the hardware implementation of PC with spintronic technology, highlighting the advantages in terms of energy cost and speed.