The Community for Technology Leaders
2010 19th International Conference on Parallel Architectures and Compilation Techniques (PACT) (2010)
Vienna, Austria
Sept. 11, 2010 to Sept. 15, 2010
ISBN: 978-1-5090-5032-1
pp: 431-442
Walid J. Ghandour , Electrical and Computer Engineering Department, American University of Beirut, Lebanon 1107 2020
Haitham Akkary , Electrical and Computer Engineering Department, American University of Beirut, Lebanon 1107 2020
Wes Masri , Electrical and Computer Engineering Department, American University of Beirut, Lebanon 1107 2020
ABSTRACT
Value prediction is a technique to increase parallelism by attempting to overcome serialization constraints caused by true data dependences. By predicting the outcome of an instruction before it executes, value prediction allows data dependent instructions to issue and execute speculatively, hence increasing parallelism when the prediction is correct. In case of a misprediction, the execution is redone with the corrected value. If the benefit from increased parallelism outweighs the misprediction recovery penalty, overall performance could be improved. Enhancing performance with value prediction therefore requires highly accurate prediction methods. Most existing general value prediction techniques are local and future outputs of an instruction are predicted based on outputs from previous executions of the same instruction.
INDEX TERMS
value prediction, computer architecture, correlation, dynamic information flow analysis, information flow strength, information theory, instruction level parallelism, program dependence analysis
CITATION
Walid J. Ghandour, Haitham Akkary, Wes Masri, "The potential of using dynamic information flow analysis in data value prediction", 2010 19th International Conference on Parallel Architectures and Compilation Techniques (PACT), vol. 00, no. , pp. 431-442, 2010, doi:
198 ms
(Ver 3.3 (11022016))