Parallel System Resource Management

有效的资源管理算法对计算机系统的性能至关重要。本项目将结合实验技术、仿真和分析性能模型，解决与并行计算机系统中有效资源管理相关的开放性问题。具体而言，该项目将研究有关高性能并行处理器分配策略、并行处理器分配与内存管理之间的相互作用、并行编程模型与评估并行资源管理算法的适当工作负载之间的相互作用以及工作负载特征对特定并行系统资源管理算法性能的影响等问题。这个项目的互动部分包括：(1)教授两门课程，一门是研究生研讨会，另一门是研究生或本科生的核心课程；(2)模仿最近计算机研究协会（CRA）的同名研讨会，主持一个关于女性学术生涯的西北研讨会；(3)参与华盛顿大学工程项目中的女性的指导和推广工作；(4)与华盛顿大学计算机科学与工程学院和学生的联合研究。

Effective resource management algorithms are critical to the performance of a computer system. This project will address open questions related to effective resource management in parallel computer systems, using a combination of experimental techniques, simulation, and analytical performance models. In particular the project will investigate questions concerning high performance parallel processor allocation policies, the interplay between parallel processor allocation and memory management, the interplay between parallel programming models and appropriate workloads for evaluating parallel resource management algorithms, and the impact of workload characteristics on the performance of particular parallel system resource management algorithms. The interactive component of this project involves: (1) teaching two courses, one a graduate seminar and one a core graduate or undergraduate course, (2) hosting a Northwest Workshop on Academic Careers for Women, modeled after the recent Computing Research Association (CRA) workshops of the same title, (3) participation in the mentoring and outreach efforts of the Women in Engineering program at the University Washington, and (4) joint research with Computer Science and Engineering faculty and students at the University of Washington.