ExpandQISE: Track 1: Collaborative Optimization and Management for Iterative and Parallel Quantum Computing

ExpandQISE：轨道 1：迭代和并行量子计算的协作优化和管理

• 批准号: 2329020
• 负责人: Ying Mao
• 金额: $ 69.06万
• 依托单位: Fordham University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-11-15 至 2026-10-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2329020&HistoricalAwards=false
• 关键词: ExpandQISE; Track; Collaborative; Optimization; Management

Non-technical Abstract: Quantum computing is advancing rapidly, pushing the boundaries of classical computing. Due to limited quantum resources on any single machine, recent research attempts to integrate multiple quantum machines and classical computing nodes into large-scale, multi-node systems. However, existing applications and management schemes need to be optimized with respect to these systems. Focusing on iterative and parallel quantum applications, this project studies collaborative optimizations and management approaches that aim to fully exploit the potential of a multi-node quantum-classical system. Building on popular open-source projects, such as Qiskit and Kubernetes, the proposed system creates an accessible entry point for researchers from diverse computer science backgrounds to explore quantum computing. Moreover, the team is committed to promoting quantum computing to underrepresented students through multi-institutional workshops and hackathons. These collaborative workforce developments and capacity-building activities will provide opportunities for students at a non-R1 institution to develop their quantum knowledge base, paving the way for a more diverse and skilled workforce in the emerging field of quantum computing.Technical Abstract: The research objective of this project is to investigate a quantum-classical system that can collaboratively optimize iterative and parallel quantum applications. The proposed system combines multiple quantum machines with classical computing nodes, operating in three modes: simulation-only, quantum-classical, and quantum-only. Specifically, it supports collaborative optimizations which utilize a circuit analyst to identify iterative and dependent operations automatically. It further generates application-specific logical and physical optimization plans to fully leverage the available quantum and classical resources within the system. Additionally, a collaborative management pipeline is developed with a visualized representation. The system considers multiple quantum and classical nodes as a unified system and optimizes input-output roundtrip loopback between quantum and classical computing components. It utilizes an online-offline combined approach to address the imbalanced executions. Furthermore, it incorporates hardware-informed optimization and management, which seeks to effectively gather the hardware’s noise, calibration, and connectivity information in an efficient way to optimize the quantum application’s fidelity. It queries the hardware to track parameters of importance to the desired quantum algorithm. Ultimately, this project provides a comprehensive approach to tackling the challenges faced in multi-node, quantum-classical systems, ensuring efficient resource utilization and seamless integration.This project is funded by The Computing and Communications Foundations (CCF) Division.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

非技术摘要：量子计算正在迅速发展，推动经典计算的边界。由于单个机器上的量子资源有限，最近的研究试图将多个量子机器和经典计算节点集成到大规模的多节点系统中。然而，现有的应用和管理方案需要针对这些系统进行优化。该项目专注于迭代和并行量子应用，研究旨在充分利用多节点量子经典系统潜力的协作优化和管理方法。基于流行的开源项目，如Qiskit和Kubernetes，拟议的系统为来自不同计算机科学背景的研究人员探索量子计算创造了一个可访问的入口点。此外，该团队致力于通过多机构研讨会和黑客马拉松向代表性不足的学生推广量子计算。这些合作性的劳动力发展和能力建设活动将为非R1机构的学生提供发展量子知识基础的机会，为新兴量子计算领域的更多样化和熟练的劳动力铺平道路。技术摘要：本项目的研究目标是研究一个量子经典系统，该系统可以协同优化迭代和并行量子应用。该系统将多个量子机与经典计算节点相结合，以三种模式运行：仅模拟，量子经典和仅量子。具体来说，它支持协同优化，利用电路分析师自动识别迭代和相关操作。它进一步生成特定于应用程序的逻辑和物理优化计划，以充分利用系统中可用的量子和经典资源。此外，协同管理管道开发与可视化表示。该系统将多个量子和经典节点视为一个统一的系统，并优化了量子和经典计算组件之间的输入-输出往返传输。它利用线上线下相结合的方法来解决不平衡的执行。此外，它还集成了硬件信息优化和管理，旨在以有效的方式有效地收集硬件的噪声、校准和连接信息，以优化量子应用的保真度。它查询硬件以跟踪对所需量子算法重要的参数。最终，该项目提供了一种全面的方法来应对多节点量子经典系统所面临的挑战，确保有效的资源利用和无缝集成。该项目由计算和通信基金会（CCF）部门资助。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。