| Andrew P. Nisbet. GAPS: Genetic algorithm optimised parallelisation. In Proceedings of the 7th International Workshop on Compilers for Parallel Computers, Linkoping, Sweden, June 1998. |
....in Chapter 3. Combining the non deterministic evolutionary principles with program optimisation may lead to a new approach to optimisation, with possibly at least as good performance as classical deterministic methods. The Genetic Algorithm Parallelisation System (GAPS) as presented by Nisbet [61], investigates just this hypothesis. GAPS is an iterative feedback guided optimising compiler which uses genetic algorithm (GA) principles to evolve good program transformations. This chapter gives an overview of the current state of GAPS, its capabilities, its limitations and its potential ....
Andrew P. Nisbet. GAPS: Genetic algorithm optimised parallelisation. In Proceedings of the 7th International Workshop on Compilers for Parallel Computers, Linkoping, Sweden, June 1998.
....space of optimisations at compile time without incurring runtime overhead. Later work could combine the approaches by including dynamic monitoring to select, at runtime, one of a number of optimisations programs that were determined (at compile time) to perform well under certain circumstances. In [6], genetic algorithms are used to create and select transformations for parallel optimisation. This work is similar in spirit to the work presented in this paper but at present generates many illegal programs which must be discarded and hence examines a much larger set of programs before finding ....
A. Nisbet, GAPS: Genetic Algorithm Optimised Parallelisation. Proc. 7th Workshop on Compilers for Parallel Computing, 1998.
No context found.
A.P. Nisbet, GAPS: Genetic Algorithm Optimised Parallelisation, 7th Workshop on Compiler for Parallel Computers, Linkøping, Sweden, June 1998.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC