Quantum Circuit Partitioning for Faster Tensor Product Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The computational complexity and time required for classical computation of tensor products increase significantly when a quantum circuit is divided into multiple sub-circuits, leading to longer overall computation times.
Innovation Solution
A method is introduced to divide a quantum circuit into sub-circuits and perform quantum computations with specific basis conversions and initial values, followed by classical computation of tensor products, optimizing the scheduling to minimize computation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a quantum circuit is divided into multiple sub-circuits for parallel execution, then computational accuracy is improved through high-fidelity small-scale quantum computers, but the processing time for classical computation of tensor products increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing basis conversion matrices and initial value vectors before the main quantum computation. These pre-computed components are then reused during tensor product calculations, avoiding redundant computations and significantly reducing the time required for classical post-processing of quantum results.
Solution Approach 2:
The patent segments the quantum circuit into multiple sub-circuits that can be executed in parallel on different quantum hardware units. Each sub-circuit is independently executed and measured, with results later combined through tensor products. This segmentation enables parallel processing while maintaining accuracy, and the patent进一步优化 by organizing the combination process into efficient sequential steps.
2Productivity
If the number of divisions of quantum circuit increases, then parallel execution capability is improved, but the computational complexity of classical computation increases on the order of O(4^K)
Solution Approach 1:
The patent pre-computes basis conversion matrices and initial value vectors for all possible division points and combinations before executing the quantum circuits. This preliminary computation allows the main execution phase to simply retrieve and multiply pre-computed values, reducing the computational complexity from O(4^K) to a much lower complexity class while maintaining full parallel execution capability.
Solution Approach 2:
The patent creates universal basis conversion matrices and initial value vectors that can be reused across multiple quantum circuit executions. These pre-computed components serve multiple purposes: they can be applied to different sub-circuits, different division points, and different quantum hardware configurations, thereby reducing the overall computational complexity while maintaining high parallel execution capability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing apparatus divides a quantum circuit to generate first and second sub-circuits. The information processing apparatus generates a plurality of combinations each including one of basis conversions to be performed at an end portion of the first sub-circuit and one of initial values to be set at a start portion of the second sub-circuit. The information processing apparatus sequentially selects a combination to be used for execution from the plurality of combinations, and causes a quantum computer to execute, in order from the selected combination, a first quantum computation including execution of the first sub-circuit and measurement corresponding to the selected combination, and a second quantum computation including initialization corresponding to the selected combination and execution of the second sub-circuit. The information processing apparatus computes tensor products based on execution results of the first and second quantum computations and a sum of the tensor products.