Picking statistically valid and early simulation points
Erez Perelman, Greg Hamerly, Brad Calder · 2003
Modern architecture research relies heavily on detailed pipeline simulation. Simulating the full execution of an industry standard benchmark can take weeks to months to complete. To address this issue we have recently proposed using Simulation Points (found by only examining basic block execution frequency profiles) to increase the efficiency and accuracy of simulation. Simulation points are a small set of execution samples that when combined represent the complete execution of the program.
Cited by 209 works. doi:10.5555/942806.943854
Authors
References
- Some methods for classification and analysis of multivariate observations
- Statistical Inference
- The SimpleScalar tool set, version 2.0
- ATOM
- X-means: Extending K-means with Efficient Estimation of the Number of Clusters
- Automatically characterizing large scale program behavior
- A Reference Bayesian Test for Nested Hypotheses and its Relationship to the Schwarz Criterion
- Basic Block Distribution Analysis to Find Periodic Behavior and Simulation Points in Applications
- Accurate low-cost methods for performance evaluation of cache memory systems
- Reducing state loss for effective trace sampling of superscalar processors