Publication: Reducing reverse-mode memory requirements by using profile-driven checkpointing
Introduction
Applications
Tools
Research Groups
Workshops
Publications
   List Publications
   Advanced Search
   Info
   Add Publications
My Account
About
Impress

Reducing reverse-mode memory requirements by using profile-driven checkpointing

- Article in a journal -
 

Area
Aerodynamics

Author(s)
Mike Fagan , Alan Carle

Published in
Future Generation Computer Systems

Year
2005

Abstract
Reverse-mode derivative calculations have favorable time cost for many problems. Unfortunately “real world” reverse-mode computations frequently experience prohibitive space costs. To mitigate this space cost, users resort to checkpointing techniques to recompute, rather than save, the necessary values. Injudicious checkpointing, however, can destroy the favorable time performance that made reverse mode attractive in the first place. Consequently, reverse-mode users must spend significant amounts of development time analyzing and developing checkpointing schemes that complement their reverse-mode computation code. In this paper, we describe a particular instance of this checkpointing problem: we were using reverse-mode code generated by ADIFOR 3.0 to compute derivatives of a large computational fluid dynamics code. Our effort labored under the additional constraint that development time was minimal (as always, it was needed yesterday). Our solution was to use profiling to narrowly focus our checkpoint analysis. This profiling approach worked well for our problem. Furthermore, the profiling idea is sufficiently general that it should work well for other problems. This paper details both our results on our specific problem and guidelines for applying the profiling technique to other checkpoint-based reverse-mode development problems.

AD Tools
ADIFOR

AD Theory and Techniques
Reverse Mode

BibTeX
@ARTICLE{
         Fagan2005Rrm,
       title = "Reducing reverse-mode memory requirements by using profile-driven checkpointing",
       journal = "Future Generation Computer Systems",
       pages = "1380--1390",
       doi = "http://dx.doi.org/10.1016/j.future.2004.11.005",
       url = "http://www.sciencedirect.com/science/article/pii/S0167739X04001815",
       author = "Mike Fagan and Alan Carle",
       keywords = "Profiling",
       abstract = "Reverse-mode derivative calculations have favorable time cost for many problems.
         Unfortunately “real world” reverse-mode computations frequently experience
         prohibitive space costs. To mitigate this space cost, users resort to checkpointing techniques to
         recompute, rather than save, the necessary values. Injudicious checkpointing, however, can destroy
         the favorable time performance that made reverse mode attractive in the first place. Consequently,
         reverse-mode users must spend significant amounts of development time analyzing and developing
         checkpointing schemes that complement their reverse-mode computation code. In this paper, we
         describe a particular instance of this checkpointing problem: we were using reverse-mode code
         generated by Adifor 3.0 to compute derivatives of a large computational fluid dynamics code. Our
         effort labored under the additional constraint that development time was minimal (as always, it was
         needed yesterday). Our solution was to use profiling to narrowly focus our checkpoint analysis. This
         profiling approach worked well for our problem. Furthermore, the profiling idea is sufficiently
         general that it should work well for other problems. This paper details both our results on our
         specific problem and guidelines for applying the profiling technique to other checkpoint-based
         reverse-mode development problems.",
       volume = "21",
       number = "8",
       year = "2005",
       ad_area = "Aerodynamics",
       ad_tools = "ADIFOR",
       ad_theotech = "Reverse Mode"
}


back
  

Contact:
autodiff.org
Username:
Password:
(lost password)