Timing constraints due to real-time graph traversal algorithms on incomplete cluster states in photonic measurement-based quantum computing

TL;DR

This paper analyzes the classical control overheads in photonic measurement-based quantum computing, focusing on real-time graph traversal algorithms for incomplete cluster states, and compares different algorithmic approaches to optimize speed and accuracy.

Contribution

It introduces and evaluates two graph traversal algorithms, highlighting their tradeoffs and implications for the photonic clock cycle in quantum computing.

Findings

Global BFS provides higher accuracy but slower execution.

Incremental traversal offers faster performance with some accuracy loss.

Overheads significantly impact the feasible clock cycle of photonic quantum systems.

Abstract

Understanding the computational overheads imposed by classical control systems on quantum computing platforms becomes critically important as these quantum machines grow in scale and complexity. In this work, we calculate the overheads imposed by the implementation of real-time graph traversal algorithms needed to find computational paths through incomplete cluster states for the implementation of one-qubit gates; a necessary requirement for a realistic implementation of photonic measurement-based quantum computing. By implementing two different algorithms, a global breadth-first search that searches the entire cluster state and an incremental version that traverses a narrow sub-section of the cluster state, we analyze the tradeoff between the accuracy of finding viable paths and the speed at which this operation can be performed, which constrains the overall photonic clock cycle of the system. We also outline the broader implications of our results for implementing classical control systems for measurement-based photonic quantum computing.