Journal
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
Volume 32, Issue 2, Pages 330-341Publisher
IEEE COMPUTER SOC
DOI: 10.1109/TPDS.2020.3018636
Keywords
Tools; Optimization; Programming; Syntactics; Semantics; Guidelines; Natural language processing; Performance tools; natural language processing; code optimization
Funding
- DOE Early Career Award [DE-SC0013700]
- National Science Foundation (NSF) [1455404, 1455733, 1525609]
- Direct For Computer & Info Scie & Enginr
- Division of Computing and Communication Foundations [1455404] Funding Source: National Science Foundation
- Direct For Computer & Info Scie & Enginr
- Division of Computing and Communication Foundations [1455733, 1525609] Funding Source: National Science Foundation
Ask authors/readers for more resources
Egeria is the first automatic synthesizer of advising tools for High-Performance Computing (HPC), which constructs a text retrieval tool based on HPC programming guides to improve program performance and provide optimization knowledge. It utilizes natural language processing techniques and HPC-specific knowledge in its multi-layered design.
This article presents Egeria, the first automatic synthesizer of advising tools for High-Performance Computing (HPC). When one provides it with some HPC programming guides as inputs, Egeria automatically constructs a text retrieval tool that can advise on what to do to improve the performance of a given program. The advising tool provides a concise list of essential rules automatically extracted from the documents and can retrieve relevant optimization knowledge for optimization questions. Egeria is built based on a distinctive multi-layered design that leverages natural language processing (NLP) techniques and extends them with HPC-specific knowledge and considerations. This article presents the design, implementation, and both quantitative and qualitative evaluation results of Egeria.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available