Ising Model Ground State Search via Bipartite Graph Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods face challenges in efficiently searching for the ground state of an Ising model due to the dense structure of interactions between spins, which hinders simultaneous stochastic processing and thus slows down the processing speed in semiconductor devices.
Innovation Solution
Representing the interaction relation of an Ising model as a complete bipartite graph, where spins from two groups are connected, allowing for simultaneous updates and efficient search for the ground state using a Markov chain Monte Carlo method.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the interaction between spins is represented as a dense structure where individual spins are adjacent to all other spins, then the Ising model accurately represents combinatorial optimization problems, but stochastic processing cannot be performed on spins simultaneously, resulting in reduced processing speed
Solution Approach 1:
The patent divides the dense spin interaction system into two separate groups (first spin group and second spin group), creating a bipartite graph structure. This segmentation allows spins within each group to be processed simultaneously through independent stochastic operations, while maintaining accurate representation of the original optimization problem through controlled interaction patterns between groups.
Solution Approach 2:
The patent transforms the traditional single-group dense interaction structure into a two-dimensional bipartite structure by introducing a group dimension. Spins are organized into first and second groups with specific interaction rules, adding a structural dimension that enables parallel processing while preserving the essential optimization characteristics of the original dense interaction model.
2Productivity
If simultaneous stochastic processing of spins is attempted in a dense interaction structure, then processing speed can be increased, but the dense structure prevents independent simultaneous updates of spins
Solution Approach 1:
By segmenting spins into two independent groups with distinct update rules, the patent enables simultaneous stochastic processing within each group. The segmentation creates operational independence that allows parallel updates without the conflicts inherent in dense single-group structures, directly achieving the goal of simultaneous processing.
3Device complexity
If a complete graph structure is used to represent all spin interactions, then the model is compact and simple, but it hinders efficient parallel processing and increases computational complexity
Solution Approach 1:
The patent replaces the complete graph structure with a segmented bipartite graph structure. While the complete graph is simpler in concept, the segmented structure maintains comparable simplicity while dramatically improving parallel processing efficiency by allowing independent stochastic updates within each group, thus resolving the contradiction between structural simplicity and processing efficiency.
Solution Approach 2:
By introducing the group dimension and creating a bipartite structure, the patent transforms the single-layer complete graph into a two-layer structure. This dimensional change preserves the essential connectivity needed for accurate optimization problem representation while enabling efficient parallel processing through independent group operations.
Data Source
AI summary
To efficiently search for a ground state of an Ising model and efficiently solve a combinatorial optimization problem. An information processing device represents an interaction relation of an Ising model as a complete bipartite graph in which N spins of a first spin group and N spins of a second spin group are connected to each other, stores an energy function in which an interaction between an i-th spin of the first spin group and a j (=i)-th spin of the second spin group is set such that the i-th spin of the first spin group and the j-th spin of the second spin group have the same value and searches for a ground state of the Ising model based on the energy function. The information processing device searches for the ground state by applying an algorithm of a simulated annealing method to the above-described energy function.