| （英） |
In fault-tolerant quantum computing, a large number of physical qubits are required to construct a single logical qubit, and a single quantum node would be able to hold only a small number of logical qubits. Thus, the idea of distributed quantum computing~(DQC) is necessary to demonstrate large-scale quantum computation using these limited-scale nodes. However, previous DQC studies have primarily focused on medium-scale nodes. The design of distributed systems on small-scale nodes has not been explored yet.
Here, we propose the benchmark suite and high-performance designs of DQC systems where each node has a single logical qubit and a few connectors. First, we evaluate the computation speed of DQC systems using quantum phase estimation algorithms as a benchmark. The results imply that there are regions where the speed of inter-node communication becomes the bottleneck. To address this issue, we propose a method to improve the effective bandwidth of inter-node communication by multiplexing nodes and communication paths. According to our simulation, the proposed method can relax the communication bottleneck and significantly improve the total runtime. Our results suggest that with appropriate multiplexing, the bottleneck of communication speed can be eliminated up to any desired level, which can drastically enhance the performance of DQC. |