Distributed quantum architecture search
arxiv(2024)
摘要
Variational quantum algorithms, inspired by neural networks, have become a
novel approach in quantum computing. However, designing efficient parameterized
quantum circuits remains a challenge. Quantum architecture search tackles this
by adjusting circuit structures along with gate parameters to automatically
discover high-performance circuit structures. In this study, we propose an
end-to-end distributed quantum architecture search framework, where we aim to
automatically design distributed quantum circuit structures for interconnected
quantum processing units with specific qubit connectivity. We devise a circuit
generation algorithm which incorporates TeleGate and TeleData methods to enable
nonlocal gate implementation across quantum processing units. While taking into
account qubit connectivity, we also incorporate qubit assignment from logical
to physical qubits within our QAS framework. A two-stage progressive
training-free strategy is employed to evaluate extensive circuit structures
without circuit training costs. Through numerical experiments on three VQE
tasks, the efficacy and efficiency of our scheme is demonstrated.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要