An essential step in designing a new computer architecture is the careful examination of different design options. At present researchers use very detailed simulators to estimate processor performance, which models every cycle of an executing program (simulating every cycle of a real program) can take weeks or months. To address this problem authors present a tool SimPoint[1] that uses k-means clustering algorithms to automatically find repetitive patterns in a programs execution. By simulating one representative of each repetitive behavior pattern, simulation time can be reduced to minutes instead of weeks for standard benchmark programs, with very little cost in terms of accuracy. This repository contain paper discussion summary (As a course work for High Perfomance Computing[2]).
[1] Hamerly, Perelman, Lau, Calder, Sherwood, "Using Machine Learning to Guide Architecture Simulation", The Journal of Machine Learning Research (Volume 7), (http://dl.acm.org/citation.cfm?id=1248547.1248559)