隨著電腦科技的飛速發展，我們擁有非常多的硬體架構選擇。然而對於系統設計者來說，如何找出擁有最佳執行效能的硬體架構，卻也越來越困難。儘管我們擁有許多的模擬器可以使用，但龐大的設計空間依然使得模擬實驗所耗的時間過長。因此，本篇論文提出一個執行效能預測的方法，藉由少量的模擬實驗就可計算出所有設計空間中的執行效能，進而減少實驗的時間。我們的預測方法是利用線性迴歸做為統計上的計算模型，無論是CPU或GPU的硬體架構都可使用。實驗的結果顯示我們不僅成功的減少實驗時間，也能維持一定的預測準確度。

[1] X. Zhang, Z. Xu, Multiprocessor Scalability Predictions Through Detailed Program Execution Analysis. Proceedings of the 9th international conference on Supercomputing, pp. 97-106. 1995.
[2] R. Plackett and J. Burman. The Design of Optimum Multifactorial Experiments. Biometrika, Vol. 33, No. 4, June 1956, pp. 305-325
[3] Arjan J.C. van Gemund. Performance Prediction of Parallel Processing Systems: The PAMELA Methodology. Proceedings of the 7th international conference on Supercomputing, pp. 318-327. 1993.
[4] Z. Xu, X. Zhang, L. Sun. Semi-Empirical Multiprocessor Performance Predictions. Journal of Parallel and Distributed Computing, Vol.39, No. 1, pp. 14-28, 1996.
[5] B. Lee ,and D. Brooks. Accurate and Efficient Regression Modeling for Microarchitectural Performance and Power Prediction. ACM SIGARCH Computer Architecture News, vol. 34, no. 5, Dec. 2006, pp. 185-194.
[6] B. Lee, J. Collins, H. Wang, D. Brooks. CPR: Composable Performance Regression for Scalable Multiprocessor Models. 41st International Symposium on Microarchitecture (Micro-41), 2008, pp. 270-281
[7] lp solve reference guide, http://lpsolve.sourceforge.net/5.5/, 2009.
[8] R. Bellman, On the approximation of curves by line segments using dynamic programming, Communications of the ACM, Volume 4, Issue 6, June 1961.
[9] R. Bellman, B. Kotkin, On the approximation of curves by line segments using dynamic programming-II, Technical Report RM-2978-PR, US Air Force Project RAND, Feb. 1962.
[10] NIST/SEMATECH e-Handbook of Statistical Methods, http://www.itl.nist.gov/div898/handbook/
[11] NVIDIA CUDA, http://www.nvidia.com/object/cuda what is.html.
[12] Simplescalar LLC: http://www.simplescalar.com/.
[13] Dhrystone Benchmark, From Wikipedia, the free encyclopedia, http://en.wikipedia.org/wiki/Dhrystone.
[14] P.A.W. Lewis, E.j.Orav, Simulation Methodology for Statistician, Operations Analysts, and Engineers Volome I, Chapman & Hall/CRC, Dec. 1988
[15] CUDA Programming Guide, http://developer.download.nvidia.com/compute/cuda/1 0/NVIDIA CUDA Programming Guide 1.0.pdf
[16] GPGPU-Sim Tutorial, http://www.ece.ubc.ca/~aamodt/gpgpu-sim/tutorial/
[17] M. Saar-Tsechansky and F. Provost. Active learning for class probability estimation and ranking. In 17th International Joint Conference on Artificial Intelligence, Aug. 2001
[18] E. Ipek, S. A. Mckee, B.R. de Supinski, R. Caruana, Efficiently exploring architectural design spaces via predictive modeling, ASPLOS’2006
[19] S. Hong, H. Kim, An analytical model for a GPU architecture with memory-leveland thread-level parallelism awareness, ISCA’09
[20] W.C. Lin, CUDA平行運算機制, financelab.nctu.edu.tw/Students/LinWeiChen/
[21] T. S. Ferguson, Linear Programming Guide, http://www.usna.edu/Users/weapsys/ avramov/Compressed%2520sensing%2520tutorial/LP.pdf
[22] Y.T. Steve Li, Shared Malik, Andrew Wolf, Efficient Microarchitecture Modeling and Path Analysis for Real-Time Software, 1995
[23] Arjan J.C. van Gemund, The PAMELA Methodology, 1993
[24] X. Zhang, Z. Xu, Multiprocessor Scalability Predictions Through Detailed Program Execution Analysis
[25] Benjamin C. Lee, Bronis R. de Supinski, Methods of Inference and Learning for Performance Modeling of Parallel Applications, ASPLOS, 2007