Back to results

Massachusetts Institute of Technology

Cilkpride : always-on visualizations for parallel programming

Abstract

dc:description.abstract

Parallel programming is an increasingly important way for programmers to squeeze more performance out of their programs. Parallelization is error-prone, however, and programmers often forget to run error checkers and performance analyzers regularly. This thesis presents Cilkpride, an IDE plug-in that uses always-on visualizations to show programmers information on on their parallel program directly inside their IDE. Cilkpride runs a race checker and program profiler every time code is changed and immediately displays output to make programmers always aware of parallelization errors and performance bottlenecks. Programmers can then react and fix these issues quickly. To evaluate the system, we asked students who had taken MIT's 6.172 class, a performance engineering course, to use Cilkpride. Students found Cilkpride useful, helping them find races and bottlenecks.

Degree

thesis:*
Department dc:contributor.department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.
Grantor dc:publisher
Massachusetts Institute of Technology
Year dc:date.issued
2017

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Chau, Genghis
Advisor dc:contributor.advisor
  • Robert C. Miller.

Subjects

dc:subject × 1

Rights

dc:rights
Statement dc:rights
  • MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.
Language dc:language.iso
eng

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/1721.1/112834
OAI identifier oai:identifier
oai:dspace.mit.edu:1721.1/112834

Chain of custody

source
Harvested from
MIT
Base URL
dspace.mit.edu/oai/request
Last updated
2026-07-22
Source record
OAI-PMH GetRecord
citation

Chau, Genghis. Cilkpride : always-on visualizations for parallel programming. Massachusetts Institute of Technology, 2017. http://hdl.handle.net/1721.1/112834