Abstract:Computationally challenging combinatorial optimization problems (COPs) play a fundamental role in various applications. To tackle COPs, many Ising machines and Quadratic Unconstrained Binary Optimization (QUBO) solvers have been proposed, which typically involve direct transformation of COPs into Ising models or equivalent QUBO forms (D-QUBO). However, when addressing COPs with inequality constraints, this D-QUBO approach introduces numerous extra auxiliary variables, resulting in a substantially larger search space, increased hardware costs, and reduced solving efficiency. In this work, we propose HyCiM, a novel hybrid computing-in-memory (CiM) based QUBO solver framework, designed to overcome aforementioned challenges. The proposed framework consists of (i) an innovative transformation method (first to our known) that converts COPs with inequality constraints into an inequality-QUBO form, thus eliminating the need of expensive auxiliary variables and associated calculations; (ii) "inequality filter", a ferroelectric FET (FeFET)-based CiM circuit that accelerates the inequality evaluation, and filters out infeasible input configurations; (iii) %When feasible solutions are detected, a FeFET-based CiM annealer that is capable of approaching global solutions of COPs via iterative QUBO computations within a simulated annealing process. The evaluation results show that HyCiM drastically narrows down the search space, eliminating $2^{100} \text{ to } 2^{2536}$ infeasible input configurations compared to the conventional D-QUBO approach.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to efficiently solve general combinatorial optimization problems (COPs) with inequality constraints. When dealing with such problems, traditional Ising machines and QUBO solvers usually need to directly convert COPs into Ising models or the equivalent unconstrained quadratic binary optimization (QUBO) form. However, when dealing with COPs with inequality constraints, this method will introduce a large number of auxiliary variables, resulting in a significant expansion of the search space, an increase in hardware cost, and a decrease in solution efficiency. To this end, the paper proposes HyCiM, a new QUBO - solving framework based on hybrid in - memory computing (CiM), aiming to overcome the above challenges. The main contributions of HyCiM include: 1. **Innovative conversion method**: A new conversion method is proposed to convert COPs with inequality constraints into inequality - QUBO form, thereby avoiding expensive auxiliary variables and their related calculations, and significantly reducing the complexity of the search space and the objective function. 2. **Inequality filter**: A CiM circuit based on ferroelectric field - effect transistors (FeFET) is designed to accelerate inequality evaluation and filter out infeasible input configurations. 3. **FeFET - based CiM annealer**: A FeFET - based CiM annealer is proposed, which can approach the global optimal solution through iterative QUBO calculations during the simulated annealing process. The evaluation results show that HyCiM significantly reduces the search space, eliminates infeasible input configurations from 2 ^ 100 to 2 ^ 2536, and reduces the search space to 2 ^ 100 feasible input configurations. This not only significantly reduces the hardware area overhead (88.06% to 99.96%), but also improves the solution efficiency, with an average success rate of 98.54%, while the traditional D - QUBO implementation is only 10.75%.

HyCiM: A Hybrid Computing-in-Memory QUBO Solver for General Combinatorial Optimization Problems with Inequality Constraints

Hybrid Optimization Method Based on Membrane Computing

A Ferroelectric Compute-in-Memory Annealer for Combinatorial Optimization Problems

A Copositive Framework for Analysis of Hybrid Ising-Classical Algorithms

Quadratic unconstrained binary optimization problem preprocessing: Theory and empirical analysis

A Compute-in-Memory Annealing Processor With Interaction Coefficient Reuse and Sparse Energy Computation for Solving Combinatorial Optimization Problems

Hybrid Quantum Annealing Decomposition Framework for Unit Commitment

FlexSpin: A CMOS Ising Machine With 256 Flexible Spin Processing Elements With 8-b Coefficients for Solving Combinatorial Optimization Problems

Efficient Digital Quadratic Unconstrained Binary Optimization Solvers for SAT Problems

Optimized QUBO formulation methods for quantum computing

Solving Combinatorial Optimization Problems on Fujitsu Digital Annealer

3-Regular 3-XORSAT Planted Solutions Benchmark of Classical and Quantum Heuristic Optimizers

COPPER: a Combinatorial Optimization Problem Solver with Processing-in-memory Architecture

30.5 A Variation-Tolerant In-eDRAM Continuous-Time Ising Machine Featuring 15-Level Coefficients and Leaked Negative-Feedback Annealing

Linearizing Binary Optimization Problems Using Variable Posets for Ising Machines

Solving Quadratic Unconstrained Binary Optimization with divide-and-conquer and quantum algorithms

Accuracy and Performance Evaluation of Quantum, Classical and Hybrid Solvers for the Max-Cut Problem

Quantum-Inspired Solvers on Mixed-Integer Linear Programming Problem

Analog Iterative Machine (AIM): using light to solve quadratic optimization problems with mixed variables

Hybrid Quantum Algorithms integrating QAOA, Penalty Dephasing and Zeno Effect for Solving Binary Optimization Problems with Multiple Constraints