Skip navigation

DSpace JSPUI

DSpace preserves and enables easy and open access to all types of digital content including text, images, moving images, mpegs and data sets

Learn More
DSpace logo
English
中文
  • Browse
    • Communities
      & Collections
    • Publication Year
    • Author
    • Title
    • Subject
    • Advisor
  • Search TDR
  • Rights Q&A
    • My Page
    • Receive email
      updates
    • Edit Profile
  1. NTU Theses and Dissertations Repository
  2. 電機資訊學院
  3. 資訊工程學系
Please use this identifier to cite or link to this item: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/97780
Title: 利用計算叢集優化 QAOA 的大規模模擬
Optimizing Large-Scale QAOA Simulation using Computing Clusters
Authors: 邱信瑋
Shin-Wei Chiu
Advisor: 洪士灝
Shih-hao Hung
Keyword: 量子近似演算法,量子計算,量子模擬,組合優化問題,平行計算,效能分析,
Quantum approximate optimization algorithm (QAOA),Quantum computing,Quantum circuit simulation,Combinatorial optimization problem,Parallel programming,Performance analysis,
Publication Year : 2025
Degree: 碩士
Abstract: 本研究開發了一款資源高效的量子電路模擬器,專為在大規模組合優化問題上執行量子近似優化算法 (QAOA) 而設計。我們的核心貢獻在於整合了狀態向量轉置 (State Vector Transposition, SVT) 技術。這項技術透過在跨節點或跨裝置傳輸前執行本地端的量子位元(Qubit)重排,能有效集中傳輸所需的狀態向量區塊。此方法不僅顯著提升了傳輸效率,更有效減輕了模擬大型量子電路所固有的記憶體需求,大幅提升了模擬性能。這種方法有助於在多樣化的分散式計算平台(包括多節點 CPU 和多設備 GPU 架構)上實現高效的擴展。實驗驗證突顯了我們提出的模擬器卓越的性能。在大規模 QAOA 電路上的基準測試表明,與現有最先進的模擬器相比,我們的模擬器實現了顯著的加速,在GPU 平台上性能提升高達 1490 倍,在純 CPU 環境中提升 28 倍。此外,我們進行全面的可擴展性實驗證實了模擬器在分散式計算環境中保持接近最佳效率的能力。這種可擴展性使我們的工作成為模擬深度 QAOA 電路的寶貴工具,有效地彌合了當前經典計算資源的能力與量子算法研究日益增長的需求之間的差距,成功縮短了經典計算資源與量子算法研究需求之間的鴻溝。最後,我們亦針對 SVT 技術進行全面效能分析,確認其在 QAOA 模擬中的最佳化效益。
This study presents a resource-efficient quantum-circuit simulator purpose-built for running the Quantum Approximate Optimization Algorithm (QAOA) on large-scale combinatorial-optimization problems. Our key contribution is the integration of a State Vector Transposition (SVT) technique. By locally reordering qubits before inter-node or inter-device communication, SVT concentrates the state-vector blocks that actually need to be transferred, dramatically improving data-transfer efficiency while easing the memory footprint inherent to large-circuit simulation. The result is a substantial performance boost and seamless scalability across heterogeneous distributed platforms, including multi-node CPU clusters and multi-device GPU systems. Extensive experiments underscore the superior performance of the simulator. On large QAOA benchmarks, it outperforms state-of-the-art simulators by up to 1,490× on GPUs and 28× on CPUs. Scalability studies further show that the simulator maintains near-optimal parallel efficiency in distributed environments, making it a powerful tool for simulating deep QAOA circuits and closing the gap between classical computing capabilities and the rapidly growing demands of quantum-algorithm research. Finally, a thorough performance analysis confirms the optimization benefits of SVT within QAOA simulations.
URI: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/97780
DOI: 10.6342/NTU202501430
Fulltext Rights: 同意授權(全球公開)
metadata.dc.date.embargo-lift: 2025-07-17
Appears in Collections:資訊工程學系

Files in This Item:
File SizeFormat 
ntu-113-2.pdf2.83 MBAdobe PDFView/Open
Show full item record


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

社群連結
聯絡資訊
10617臺北市大安區羅斯福路四段1號
No.1 Sec.4, Roosevelt Rd., Taipei, Taiwan, R.O.C. 106
Tel: (02)33662353
Email: ntuetds@ntu.edu.tw
意見箱
相關連結
館藏目錄
國內圖書館整合查詢 MetaCat
臺大學術典藏 NTU Scholars
臺大圖書館數位典藏館
本站聲明
© NTU Library All Rights Reserved