Design and Implementation of a Context-Sensitive, Flow-Sensitive Activity Analysis Algorithm for Automatic Differentiation

TitleDesign and Implementation of a Context-Sensitive, Flow-Sensitive Activity Analysis Algorithm for Automatic Differentiation
Publication TypeConference Paper
Year of Publication2007
AuthorsShin, J, Malusare, P, Hovland, PD
Conference Name5th International Conference on Automatic Differentiation (AD 2008)
PublisherLect. Notes Comput. Sci. Eng.
Conference Location Bonn, Germany
Other NumbersANL/MCS-P1465-1207
AbstractAutomatic differentiation (AD) has been expanding its role in scientific computing. While several AD tools have been actively developed and used, a wide range of problems remain to be solved. Activity analysis allows AD tools to generate derivative code for fewer variables, leading to a faster run time of the output code. This paper describes the first context-sensitive, flow-sensitive (CSFS) activity analysis, which is developed by extending an existing context-sensitive, flow-insensitive (CSFI) activity analysis. Our experiments with eight benchmarks including the MIT General Circulation Model show that the new CSFS activity analysis is more than 26 times slower but reduces 8 and 1 overestimations for two out of the eight benchmarks compared with the existing CSFI activity analysis implementation.
PDFhttp://www.mcs.anl.gov/papers/P1465.pdf