Title :
Better Than Optimal: Fast Identification of Custom Instruction Candidates
Author :
Reddington, J. ; Gutin, G. ; Johnstone, A. ; Scott, E. ; Yeo, A.
Author_Institution :
Dept. of Comput. Sci., R. Holloway Univ. of London, Egham, UK
Abstract :
Asymptotically optimal algorithms do not always yield the fastest practical algorithm on realistic cases. We examine Gutin etal.´s recently published optimal algorithm for enumerating the set of convex subgraphs under input/output constraints with application to custom instruction identification. We show that (i) suppressing some of the machinery in their algorithm results in a sub-optimal algorithm which is significantly faster in practice on real-world examples and that (ii) the constants of proportionality in the running time for both optimal and sub-optimal versions can be significantly improved by using additional output set filtering constraints.
Keywords :
convex programming; directed graphs; instruction sets; program processors; convex subgraph; custom instruction identification; directed acyclic graph; filtering constraint; input/output constraint; instruction set; optimal algorithm; processor; Application software; Computer aided instruction; Computer science; Design automation; Filtering algorithms; Graph theory; Humans; Instruction sets; Machinery; Programming profession; candidate instruction enumeration; convex; convex set; instruction set extension;
Conference_Titel :
Computational Science and Engineering, 2009. CSE '09. International Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
978-1-4244-5334-4
Electronic_ISBN :
978-0-7695-3823-5
DOI :
10.1109/CSE.2009.167