Date of Award

2018

Document Type

Thesis

Degree Name

Master of Science in Engineering (MSE)

Department

Electrical and Computer Engineering

Committee Chair

Aleksandar Milenkovic

Committee Member

Rhonda Kay Gaede

Committee Member

B. Earl Wells

Subject(s)

Computer systems--Evaluation, Computer systems--Performance--Testing

Abstract

Computer engineers in both academia and industry rely on a standardized set of benchmarks to quantitatively evaluate the performance of modern computer systems and research prototypes. The SPEC CPU2017 benchmark suites are the most recent incarnation of standard benchmarks designed to stress a system’s processor, memory subsystem, and compiler. This thesis describes the results of measurement-based studies focusing on performance and energy-efficiency of modern Intel processors using SPEC CPU2017. The studies utilize SPEC CPU2017 run utilities as well as modern Linux tools for profiling that interface on-chip performance monitoring units. The thesis encompasses the following aspects of performance evaluation: (a) top-view characterization of individual benchmarks; (b) analysis of scalability in the context of speed and throughput metrics, while varying the number of threads for speed benchmarks and copies for rate benchmarks, respectively; (c) analysis using Intel’s Top-down Microarchitectural Analysis Method, (d) comparative performance study of different computers with Intel’s Core i7 and Xeon processors, and (e) analysis of performance impact of hardware prefetching in modern processors.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.