计算机科学 ›› 2026, Vol. 53 ›› Issue (6A): 250600165-5.doi: 10.11896/jsjkx.250600165
王德志1, 程琨2
WANG Dezhi1, CHENG Kun2
摘要: 聚焦于超导量子处理器与异构计算系统融合背景下的调度优化问题,研究了量子与经典任务在异构环境中的协同调度机制,提出一套面向混合任务图的统一调度建模方法,构建了基于最小通信代价与量子窗口预测的联合调度策略,并验证了该方法在任务延迟控制与资源利用效率方面的改进潜力。分析协同模型在高复杂度计算场景下的稳定性与扩展性,探讨异构调度中任务划分机制与系统约束适配能力。该成果对推动量子计算在并行计算系统中的工程落地及提升异构系统资源调度的智能化水平具有实际价值与理论意义。
中图分类号:
| [1] LIU Y,XU J F,XU H C,et al.Research on the development status and trends of high-performance computing applications based on supercomputers[J].Frontiers of Data and Computing,2025,7(2):68-85. [2] LI C Y,CUI G L,SHEN X J.Challenges and new opportunities for electronic structure theory[J].Scientia Sinica Chimica,2025,55(3):550-564. [3] LI W Z,WANG W.Fast SVD algorithm based on domestic he-terogeneous computing platform and its application in ocean data assimilation[J].Frontiers of Data and Computing,2024,6(1):35-45. [4] XU S,ZHANG B H,LIU Q,et al.eMD:A large-scale molecular dynamics simulation software based on heterogeneous computing[J].Frontiers of Data and Computing,2024,6(1):21-34. [5] LYU Q W,CHEN Z Q,ZHANG X Y,et al.Innovations in computer architecture under the von Neumann bottleneck[J].Application of Electronic Technique,2023,49(11):28-34. |
|
||