Abstract
A method for estimating the power consumption of an algorithm is presented. The estimation can be performed both from the C program and from the assembly code. It relies on a power model for the targeted processor. Without compilation, several targets can be compared at the C-level in order to rapidly explore the design space. The estimation can be refined afterwards at the assembly level to allow further code optimizations. The power model of the Texas Instrument TMS320C6201 is presented as a case study. Estimations are performed on real-life digital signal processing applications with average errors of 4.2% at the C-level, and1.8% at the assembly level.
Cite
CITATION STYLE
Senn, E., Julien, N., Laurent, J., & Martin, E. (2002). Power consumption estimation of a C program for data-intensive applications. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2451, pp. 332–341). Springer Verlag. https://doi.org/10.1007/3-540-45716-x_33
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
