Quantum System Selection via Coupling Map Topology Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for selecting quantum computing systems for executing pre-trained quantum machine learning models rely on simplistic heuristics, often leading to the selection of inappropriate systems due to neglect of qubit connection topology, resulting in suboptimal or failed executions.
Innovation Solution
A system that compares coupling maps of candidate quantum computers to the map on which the model was trained, using subgraph matching metrics like Gromov-Hausdorff distance or Manhattan distance to identify topologically matching systems, and adjusts the model by removing logical qubits or gates based on variable importance scores to ensure compatibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If quantum system selection is based on simplistic heuristics like qubit count comparison, then the selection process is simple and fast, but the selected quantum system is often inappropriate for the model
Solution Approach 1:
The patent changes the selection parameter from simple qubit count to coupling map topology comparison. By representing quantum systems as graphs with qubits as nodes and connections as edges, and comparing topological properties (degree sequences, adjacency matrices, spectral features), the system achieves more accurate matching while maintaining computational efficiency through automated graph isomorphism algorithms.
Solution Approach 2:
The patent replaces manual or heuristic-based system selection with an automated computational approach using graph theory and machine learning. The system automatically computes coupling map similarities, ranks candidate systems, and selects the best match without human intervention, substituting simplistic rules with sophisticated topological analysis.
2Productivity
If quantum system selection compares only the number of qubits, then the selection process is computationally efficient, but it neglects qubit connection topology leading to suboptimal selections
Solution Approach 1:
The patent segments the quantum system comparison into distinct topological features: degree sequences (qubit connectivity patterns), adjacency matrices (connection structure), and spectral features (eigenvalues of the coupling matrix). By analyzing each segment separately and combining results, the system achieves comprehensive topological comparison while maintaining computational efficiency.
Solution Approach 2:
The patent creates a universal coupling map representation that can represent any quantum system topology. The graph-based coupling map framework is universally applicable to different quantum architectures (superconducting, trapped ion, photonic) and enables consistent comparison across diverse systems using the same topological metrics.
3Adaptability or versatility
If the quantum computing system's coupling map does not match the training system's coupling map, then the model cannot be executed without modification, but model adjustment increases complexity
Solution Approach 1:
The patent performs preliminary coupling map comparison before model execution to identify compatible systems. By pre-assessing topological compatibility and ranking candidate systems, the method avoids the need for model adjustment when a matching system is available, thereby preventing the introduction of additional complexity.
Solution Approach 2:
The patent creates a copy of the training system's coupling map and compares it against candidate system coupling maps. This copying approach enables direct topological matching without modifying the original model, preserving model integrity while finding suitable execution platforms.
Data Source
AI summary
Systems/techniques that facilitate quantum system selection via coupling map comparison are provided. In various embodiments, a system can access a quantum machine learning (QML) model. In various aspects, the system can identify, from a set of quantum computing systems, a quantum computing system for the QML model, based on a comparison between a first coupling map of the quantum computing system and a second coupling map on which the QML model was trained. If the second coupling map topologically matches the first coupling map or topologically matches a subgraph of the first coupling map, the system can execute the QML model on the quantum computing system. Otherwise, the system can adjust the QML model and can accordingly execute the adjusted QML model on the quantum computing system.