没有合适的资源?快使用搜索试试~ 我知道了~
特定应用的内存子系统基准测试(计算机硕士毕业论文英文参考资料).pdf
1 下载量 64 浏览量
2023-08-22
12:22:18
上传
评论
收藏 3.31MB PDF 举报
温馨提示
试读
110页
应用程序性能通常取决于实现的内存带宽。根据指令组合和顺序、工作集大小和访问模式的特定组合,实现的内存带宽差异很大。实现良好的应用程序性能取决于在给定应用程序的约束内优化这些特征。由于缺乏有关微小变化对性能的影响的信息,因此此任务很复杂。某些信息由基准测试提供,但大多数内存基准仅限于简单的访问模式,这些模式不能代表实际应用程序中的模式。 本论文介绍了AdaptMemBench,这是一个可配置的基准框架,旨在探索从应用程序中提取的计算内核的性能能力。AdaptMemBench提供了一个框架来模拟特定于应用程序的内存访问模式。一组模板管理标准计时和测量任务。构建系统适应多面体模型,使框架为潜在的代码优化提供了方便的测试平台。 AdaptMemBench支持实验结果的可重复性,并促进结果的共享。鉴于基准测试中的微小变化会对性能产生很大影响,因此通用框架会隔离代码的测量部分。这简化了重新运行实验和移植到新系统的过程。AdaptMemBench 的优势通过一系列常见计算内核的案例研究来展示,包括:流模式、多维模板和稀疏矩阵操作。
资源推荐
资源详情
资源评论
APPLICATION-SPECIFIC MEMORY SUBSYSTEM
BENCHMARKING
by
Mahesh Lakshminarasimhan
A thesis
submitted in partial fulfillment
of the requirements for the degree of
Master of Science in Computer Science
Boise State University
May 2019
© 2019
Mahesh Lakshminarasimhan
ALL RIGHTS RESERVED
BOISE STATE UNIVERSITY GRADUATE COLLEGE
DEFENSE COMMITTEE AND FINAL READING APPROVALS
of the thesis submitted by
Mahesh Lakshminarasimhan
Thesis Title: Application-Specific Memory Subsystem Benchmarking
Date of Final Oral Examination: 5th March 2019
The following individuals read and discussed the thesis submitted by student Mahesh
Lakshminarasimhan, and they evaluated the presentation and response to questions
during the final oral examination. They found that the student passed the final oral
examination.
Catherine Olschanowsky, Ph.D. Chair, Supervisory Committee
Steven Cutchin, Ph.D. Member, Supervisory Committee
Hoda Mehrpouyan, Ph.D. Member, Supervisory Committee
The final reading approval of the thesis was granted by Catherine Olschanowsky,
Ph.D., Chair of the Supervisory Committee. The thesis was approved by the Graduate
College.
ACKNOWLEDGMENTS
I am grateful to my graduate advisor Dr. Cathie Olschanowsky for her support
and encouragement throughout the course of my Masters degree. She has taught
me most of what I know about research and has helped me sharpen my technical
writing and presentation skills. I am indebted to her for connecting me with great
opportunities and providing invaluable guidance in shaping my career. I also thank
her for funding me with graduate research assistantship supporting my graduate
studies.
I am thankful to my supervisory committee members Dr. Steven Cutchin and
Dr. Hoda Mehrpouyan for their constructive feedback in the successful completion of
my thesis. The suggestions from Dr. Cutchin for data visualization have been very
valuable.
The current and former members from the ADaPT Data Flow Optimizations lab
have been very good friends and collaborators. I was fortunate to be supported by
Travis Adsitt, the undergraduate researcher, in developing the late analysis tool and
the benchmark framework. I greatly benefited from the productive discussions I have
had with Eddie Davis.
I would like to thank the staff members of OIT Research Computing, specifically
Kelly Bryne and Tyler Bevan for their support in addressing the technical issues that
arose with the R2 HPC cluster. I appreciate the CS department staff for helping me
with various administrative tasks. I also thank Ben Peterson for his assistance with
the Onyx cluster and the workstations in the research neighborhood.
iv
This research utilized the high performance computing support of the R2 com-
pute cluster (DOI: 10.18122/B2S41H) provided by Boise State University’s Research
Computing Department. A part of thesis was funded by the U.S. National Science
Foundation (NSF) award #1849463.
v
剩余109页未读,继续阅读
资源评论
小兔子平安
- 粉丝: 209
- 资源: 1940
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功