Abstract | ||
---|---|---|
Iterative compiler optimization has been shown to out- perform static approaches. This, however, is at the cost of large numbers of evaluations of the program. This paper de- velops a new methodology to reduce this number and hence speed up iterative optimization. It uses predictive modelling from the domain of machine learning to automatically focus search on those areas likely to give greatest performance. This approach is independent of search algorithm, search space or compiler infrastructure and scales gracefully with the compiler optimization space size. Off-line, a training set of programs is iteratively evaluated and the shape of the spaces and program features are modelled. These models are learnt and used to focus the iterative optimization of a new program. We evaluate two learnt models, an indepen- dent and Markov model, and evaluate their worth on two embedded platforms, the Texas Instrument C6713 and the AMD Au1500. We show that such learnt models can speed up iterative search on large spaces by an order of magni- tude. This translates into an average speedup of 1.22 on the TI C6713 and 1.27 on the AMD Au1500 in just 2 evalua- tions. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1109/CGO.2006.37 | Symposium on Code Generation and Optimization |
Keywords | Field | DocType |
new program,compiler infrastructure,iterative compiler optimization,learnt model,focus iterative optimization,search algorithm,machine learning,search space,compiler optimization space size,iterative search,iterative optimization,amd au1500,shape,markov model,predictive modelling,predictive models,iterative methods,learning artificial intelligence,compiler optimization | Search algorithm,Iterative search,Iterative method,Markov model,Computer science,Theoretical computer science,Compiler,Optimizing compiler,Artificial intelligence,Predictive modelling,Machine learning,Speedup | Conference |
ISSN | ISBN | Citations |
2164-2397 | 0-7695-2499-0 | 180 |
PageRank | References | Authors |
7.23 | 10 | 9 |
Name | Order | Citations | PageRank |
---|---|---|---|
Felix V. Agakov | 1 | 442 | 34.22 |
Edwin V. Bonilla | 2 | 1008 | 53.32 |
John Cavazos | 3 | 584 | 26.93 |
Björn Franke | 4 | 744 | 47.31 |
Grigori Fursin | 5 | 958 | 50.76 |
M. F. P. O'Boyle | 6 | 318 | 17.11 |
John Thomson | 7 | 505 | 23.11 |
marc toussaint | 8 | 1299 | 97.23 |
Christopher K. I. Williams | 9 | 6807 | 631.16 |