Confidence estimation for speculation control
Dirk C. Grunwald, Artur W. Klauser, Srilatha Manne et al. · 1998
Modern processors improve instruction level parallelism by specu-lation. The outcome of data and control decisions is predicted, and the operations are speculatively executed and only committed if the original predictions were correct. There are a number of other ways that processor resources could be used, such as threading or eager execution. As the use of speculation increases, we believe more processors will need some form of speculation control to balance the benefits of speculation against other possible activities. Confidence estimation is one technique that can be exploited by architects for speculation control. In this paper, we introduce per-formance metrics to compare confidence estimation mechanisms, and argue that these metrics are appropriate for speculation con-trol. We compare a number of confidence estimation mechanisms, focusing on mechanisms that have a small implementation cost and gain benefit by exploiting characteristics of branch predictors, such as clustering of mispredicted branches. We compare the performance of the different confidence esti-mation methods using detailed pipeline simulations. Using these simulations, we show how to improve some confidence estimators, providing better insight for future investigations comparing and ap-plying confidence estimators. 1
Cited by 148 works. doi:10.1145/279358.279376
Authors
References
- A study of branch prediction strategies
- Simultaneous multithreading
- Evaluating future microprocessors : The SimpleScalar tool set
- Probability, Statistics, and Queueing Theory with Computer Science Applications
- Pipeline Gating: Speculation Control For Energy Reduction
- Alternative implementations of two-level adaptive branch prediction
- Assigning confidence to conditional branch predictions
- ProfileMe: hardware support for instruction-level profiling on out-of-order processors
- Disjoint eager execution: an optimal form of speculative execution
- The Statistical Precision of Medical Screening Procedures: Application to Polygraph and AIDS Antibodies Test Data