| Karel Driesen and Urs Holzle. Improving Indirect Branch Prediction With Source- and Arity-based Classification and Cascaded Prediction. Technical report, UC-Santa Barbara, March 1998. |
.... techniques [2, 13, 15, 18] The indirect branch predictors can benefit from better replacement algorithms for the target addresses stored in the predictor table [1, 6] Any of the predictors presented in this paper may be used as a component of a hybrid predictor [14] or a cascaded predictor [7]. 5.1 Methodology The results presented in this paper are for the eight SPECint95 benchmarks. We also present the results for eight other commonly used programs. These non SPEC benchmarks are: version 4.0 of GNU Chess (chess) version 1.10 of the GNU C implementation of the troff document ....
K. Driesen and U. Holzle, "Improving indirect branch prediction with source- and arity-based classification and cascaded prediction," Technical Report TRCS9807, Department of Computer Science, University of California, Santa Barbara, March 1998.
....indirect branches, conditional branches or calls returns) Therefore, they allowed different types of path based correlation to be exploited. Their simulation results clearly showed the dependence of indirect branch predictability on the type of path based correlation. Recently, Driesen and Holzle [18] performed another study where they examined the predictability of indirect branches using source code and opcode based classification, as well as filtering. Although the prediction accuracy did not justify using source code or instruction opcodes, the filtering scheme improved performance by ....
K. Driesen and U. Holzle. Improving Indirect Branch Prediction with Source- and Arity-based Classification and Cascaded Prediction. Technical Report TRCS98-07, University of California, Santa Barbara, March 1998.
No context found.
Karel Driesen and Urs Hlzle. Improving Indirect Branch Prediction With Source- and Aritybased Classification and Cascaded Prediction. Technical Report TRCS98-07, Computer Science Department, University of California, Santa Barbara, 15 March 1998 (http:// www.cs.ucsb.edu/oocsb/papers/TRCS98-07.html)
No context found.
Karel Driesen and Urs Holzle. Improving Indirect Branch Prediction With Source- and Arity-based Classification and Cascaded Prediction. Technical report, UC-Santa Barbara, March 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