簡易檢索 / 詳目顯示

回結果列表
研究生： 陳泰廷
Chen, Tai-Ting
論文名稱： 基於學習技巧能考量應用相依功耗之熱感知平面規劃法
Thermal-Aware Floorplanning Methodology Considering Application Dependent Power Based on Learning Techniques
指導教授： 林家民
Lin, Jai-Ming
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 電機工程學系
Department of Electrical Engineering
論文出版年： 2018
畢業學年度： 106
語文別： 英文
論文頁數： 57
中文關鍵詞： 平面規劃 、熱效應 、固定框架
外文關鍵詞： floorplanning, thermal effect, fixed-outline
相關次數： 點閱：134  下載：1 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 近年來，隨著晶片尺寸不斷微縮，功率不斷升高，晶片內單位面積的功率迅速增加，使得晶片內的溫度急遽升高。然而，過高的溫度或溫度不均會造成晶片的性能與可靠度下降，因此熱效應已成為晶片設計或實體設計期間不可忽視的重要議題。為了能夠準確地估測溫度，必須先確定模組位置，這使得高效率且具有熱感知的平面規劃器顯得格外重要。
    儘管過去已有許多著作針對熱感知平面規劃演算法進行研究與探討，其多半採用模擬退火法處理此問題，然而此方法通常相對耗時，且品質不穩定。因此本論文首先提出了一個利用數學解析模型，能有效率並滿足固定框架限制之熱感知平面規劃法，其能在微幅增加繞線長度的代價下大幅降低晶片內溫度。此外，大多現有的熱感知平面規劃器僅基於一筆代表性功率資料進行溫度的估測，但是由於晶片中的模組的功率會隨著不同的應用程序和時間變化，故只考量靜態功率的結果，並不足以滿足實際的狀況。因此，為了能夠進一步考慮應用相依的動態功耗變化，我們利用機器學習技術並且配合前述之快速熱感知平面規劃器，提出了一個能夠考量動態功耗的平面規劃設計流程。實驗結果顯示，本論文提出的方法，能夠考量動態的功耗，並且在不劣化過多線長的前提下，有效地避免熱點生成。更重要的是，我們的運行時間是相當快速的，並且能夠滿足固定框架限制。

    As semiconductor feature size continues shrinking, power density per area grows rapidly, which makes on-chip temperature higher than ever. High temperature or temperature non-uniformity has become a serious threat to performance and reliability of high-performance chips. Hence, thermal effect becomes a non-ignorable issue to circuit design or physical design. However, in order to estimate temperature accurately, the locations of modules have to be determined, which makes an efficient and effective thermal-aware floorplanning play a more important role.
    Although there exist some works focusing on thermal-aware floorplanning, their methods are usually based on the simulated annealing algorithm, which is very time consuming and the solution quality is not stable. Besides, most existing thermal-aware floorplanners estimate temperature based on a single power profile. But the power consumption of modules will change over time and over different applications. The results only based on static power are not enough to meet realistic cases. Hence, this thesis proposes to use analytical models to develop a fast thermal-aware fixed-outline floorplanner. Further, in order to consider the application dependent power, we also develop a design methodology which combines learning techniques as well as our thermal-aware floorplaner to find a solution that can satisfy power variation. Experimental results demonstrate that our methodology can generate a better floorplan in a design with the application dependent power. More importantly, our runtime is quite fast.

    Table of Contents 摘要 I Abstract II 誌謝 IV Table of Contents V List of Tables VII List of Figures VIII Chapter 1 Introduction 1 1.1 Previous Works 3 1.1.1 Temperature Analysis Approaches 3 1.1.2 Floorplanning Considering Thermal Issue 3 1.1.3 Floorplanning Considering Application Dependent Power 6 1.2 Our Contributions 6 1.2.1 Thermal-Aware Floorplanning Methodology 7 1.2.2 Generating Floorplans Considering Application Dependent Power based on Learning Techniques 8 1.3 Thesis Organization 9 Chapter 2 Problem Formulation 10 Chapter 3 Thermal-Aware Floorplanning Methodology 11 3.1 Preliminary 11 3.1.1 Thermal-Aware Analytical Floorplanning Framework 11 3.1.2 SAINT: 3D Fixed-outline Floorplanner 13 3.2 Overview of Our Methodology 15 3.3 Thermal-Aware Multi-Level Framework 17 3.3.1 Thermal-Aware Clustering 19 3.3.2 Balance of Different Forces in Non-linear Model 19 3.3.3 Thermal Force Modulation 21 Chapter 4 Generating Floorplans Considering Application Dependent Power based on Learning Techniques 25 4.1 Overview of Proposed Flow 25 4.2 Model-fitting Framework 26 4.2.1 Preparing Training Samples 26 4.2.2 Predictor Features and Feature Selection 27 4.2.3 Support Vector Regression 29 4.3 Predictor-Guided Optimal Floorplan Search 31 4.3.1 Generating Hotspot Candidates 31 4.3.2 Searching an Optimal Floorplan 33 Chapter 5 Experimental Results 35 5.1 Thermal-Aware Floorplanning Methodology 36 5.1.1 Comparison with Wirelength-Driven and Flat Approach 36 5.1.2 Comparison with Corblivar [11] 42 5.2 Generating Floorplans Considering Application Dependent Power based on Learning Techniques 47 5.2.1 Experiments of Temperature-Cost Model 48 5.2.2 Comparison with Different Clustering Algorithms 49 5.2.3 Comparison with Floorplans using different Power Profiles 50 Chapter 6 Conclusion 54 Bibliography 55

    Bibliography
    [1] G. Chen and S. Sapatnekar, “Partition-driven standard cell thermal placement,” in Proc. ISPD, pp. 75-80, April 2003.
    [2] B. Goplen and S. Sapatnekar, “Efficient thermal placement of standard cells in 3D ICs using a force directed approach,” in Proc. ICCAD, pp. 86-89, Nov. 2003.
    [3] C. H. Tsai and S. M. Kang, “Cell-level placement for improving substrate thermal distribution,” IEEE TCAD, vol. 19, no. 2, pp. 253- 266, Feb. 2000.
    [4] P.-Y. Huang and Y.-M. Lee, “Full-chip thermal analysis for the early design stage via generalized integral transforms,” IEEE TCAD, vol. 17, no. 5, pp. 613-626, May. 2009.
    [5] B. Wang and P. Mazumder, “Accelerated chip-level thermal analysis using multilayer Greens function,” IEEE TCAD, vol. 26, no. 2, pp. 325-344, Feb. 2007.
    [6] J. Cong, G-J. Luo, J Wei, and Y. Zhang, “Thermal-aware 3D IC placement via transformation,” in Proc. ASP-DAC, pp. 780-785, Jan. 2007.
    [7] B. Goplen and S. Sapatnekar, “Efficient thermal placement of standard cells in 3D ICs using a force directed approach,” in Proc. ICCAD, pp. 86-89, Nov. 2003.
    [8] P.-Y. Huang and Y.-M. Lee, “Full-chip thermal analysis for the early design stage via generalized integral transforms,” IEEE TCAD, vol. 17, no. 5, pp. 613-626, May 2009.
    [9] W.-L. Hung, G. M. Link, Y. Xie, N. Vijaykrishnan, and M. J. Irwin, “Interconnect and thermal-aware floorplanning for 3D microprocessors,” in Proc. ISQED, pp. 98-104, March 2006.
    [10] A. B. Kahng, S. M. Kang, W. Li, and B. Liu, “Analytical thermal placement for VLSI lifetime improvement and minimum performance variation,” in Proc. ICCAD, pp. 71-77, Oct. 2007.
    [11] J. Knechtel, E. Young, J. Lienig, “Planning massive interconnects in 3-d chips,” IEEE TCAD, vol. 34, no. 11, pp. 1808-1821, Nov. 2015.
    [12] J. Li and H. Miyashita, “Thermal-aware placement based on FM partition scheme and force-directed heuristic,” IEICE Trans. Fundam. Electron. Commun. Computet., vol. 89, no. 4, pp. 989-995, Apr. 2006.
    [13] X. Li, Y. Ma, and X. Hong, “A novel thermal optimization flow using incremental floorplanning for 3D ICs,” in Proc. ASP-DAC, pp. 347-352, Jan. 2009.
    [14] I. Arnaldo, J.L. Risco-Martín, J.L. Ayala, J.I. Hidalgo, "Power profiling-guided floorplanner for 3D multi-processor systems-on-chip", IET Circuits Devices & Systems, vol. 6, pp. 322-329, Sept. 2012.
    [15] L. Singhal, O. Sejong, E. Bozorgzadeh: ‘Statistical power profile correlation for realistic thermal estimation’. Proc. ASP-DAC, pp. 67–70, April. 2008.
    [16] J.-M. Lin, P.-Y. Chiu, and Y.-F. Chang, “SAINT: handling module folding and alignment in fixed-outline floorplans for 3D ICs,” in Proc. ICCAD, Nov. 2016.
    [17] S. Aroonsantidecha, S. S.-Y. Liu, C.-Y. Chin, and H.-M. Chen, “Fast thermal aware placement with accurate thermal analysis based on green function,” in Proc. ASP-DAC, pp. 425-430, Feb. 2012.
    [18] G. Luo, Y. Shi, and Jason Cong, “An analytical placement framework for 3-D ICs and its extension on thermal awareness,” IEEE TCAD, vol. 32, no. 4, pp. 510-523, April 2013.
    [19] B. Obermeier and F. M. Johannes, “Temperature-aware global placement,” in Proc. ASP-DAC, pp. 143-148, Jan. 2004.
    [20] C. H. Tsai and S. M. Kang, “Cell-level placement for improving substrate thermal distribution,” IEEE TCAD, vol. 19, no. 2, pp. 253-266, Feb. 2000.
    [21] B. Wang and P. Mazumder, “Accelerated chip-level thermal analysis using multilayer Greens function,” IEEE TCAD, vol. 26, no. 2, pp. 325-344, Feb. 2007.
    [22] L. Xiao, S. Sinha, J. Xu, and E. Young, “Fixed-outline thermal-aware 3d floorplanning,” in Proc. ASP-DAC, pp. 561-567, Jan. 2010.
    [23] Y. Zhan and S. S. Sapatnekar, “High efficiency green functionbased thermal simulation algorithms,” IEEE TCAD, vol. 26, no. 9, pp. 1661-1675, Sep. 2007.
    [24] P. Zhou, Y. Ma, Z. Li, R. Dick, L. Shang, H. Zhou, X. Hong, and Q. Zhou, “3D-STAF: scalable temperature and leakage aware floorplanning for three-dimensional integrated circuits,” in Proc. ICCAD, pp. 590-597, Nov. 2007.
    [25] A. Ziabari, J.-H. Park, E. Ardestani, J. Renau, S.-M. Kang, and A. Shakouri, “Power Blurring: fast static and transient thermal analysis method for packaged integrated circuits and power devices,” IEEE TVLSI, vol. 22, no. 11, pp. 2366-2379, Nov 2014.
    [26] J.-M. Lin, Y.-F. Chang, “Thermal-Aware Analytical Floorplanning Framework Based On Power Blurring Method” thesis of department of electrical engineering national Cheng Kung university, 2017.
    [27] C. Alpert, A. Kahng, G.-J. Nam, S. Reda, and P. G. Villarrubia, ”Asemipersistent clustering technique for VLSI circuit placement,” in Proceedings of ACM International Symposium on Physical Design, 2005, pp, 200-207.
    [28] HARTIGAN, J. A. 1975. Clustering Algorithms. John Wiley and Sons, Inc., New York, NY.
    [29] Toshie Yamashita , Keizo Yamashita , Ryotaro Kamimura (2007) A Stepwise AIC Method for Variable Selection in Linear Regression, Communications in Statistics—Theory and Methods. 2007;36(13):2395–2403.
    [30] C.Chang,“Libsvm: A library for support vector machines,” ACM Transaction of Intelligent System and Technology (TIST), vol. 2, 2011.
    [31] V. N. Vapnik, The Nature of Statistical Learning Theory. New York: Springer, 1995.
    [32] G. Digele, S. Lindenkreuz, and E. Kasper, “Fully coupled dynamic electro-thermal simulation,” IEEE TVLSI., Sept. 1997.
    [33] W. Huang, M. R. Stan, K. Skadron, K. Sankaranarayanan, S. Ghosh, and S. Velusamy, "Compact Thermal Modeling for Temperature-Aware Design." in Proc. DAC, pp. 878-883, Jun. 2004.

    無法下載圖示 校內：2023-07-26公開
    校外：不公開
    電子論文尚未授權公開，紙本請查館藏目錄
    QR CODE