臺灣博碩士論文加值系統

一般來說骨牌電路具有較高操作速度的優點，因此被廣泛使用在許多高性能的電路中，例如CPU及DSP晶片。雖然骨牌電路較傳統的CMOS電路有較高的執行速度，然而，雖然相較於標準CMOS電路而言，骨牌電路有運算快速的特點，然而當今市面上卻沒有專為骨牌電路設計的電路合成軟體。因此，利用傳統的電路合成流程對骨牌電路進行合成，可能無法將骨牌電路高速運算的優點完全發揮。本篇論文以此為出發點，利用各種方法，使骨牌電路實作的電路速度加以提升。
研究方向主要可以分成幾個部分，第一個部分利用修改電路的結構，將原始電路進行node(gate)合併的動作，再將合併後的結果和合成工具整合計算來達到提升速度的功能。第二個部分是利用output phase assignment將電路中並聯的邏輯閘比例提升，藉此增快骨牌電路骨牌電路的放電速度。第三部分是利用建立移除Domino cell的Foot NMOS模型，模擬並觀測Domino circuit在移除Foot NMOS後的速度提升。另外，本篇論文也針對Pin assignment進行研究，利用Pin assignment的方式來進行最後的優化，可惜未能在實驗中呈現出良好的效果，因此未列入我們最後的實驗結果中。
這項研究和工業界合作，因此所使用到的各種軟硬體工具，都和當今標準的design flow整合，證明此項研究提出的方法確實可以應用在業界的設計當中。在實際效能改進的數據上，平均而言，我們的方法相對於傳統靜態0.13um製程的CMOS電路設計可提昇47.01%的電路效能。

Domino logic is widely used in today's high performance circuit design. Although domino logic has high speed property than static CMOS logic, commercial tools do not support any special synthesis methods for domino logic. Therefore domino logic synthesis might not have good timing results when using traditional synthesis flow. In this paper, we have purposed following several methods: partially collapse, output phase assignment for timing improvement, remove foot NMOS, and pin assignment to improve circuit timing, but pin assignment does not have good results in our experiments, so we do not use in our final experiments. The other three methods presented here are combined with domino circuit traditional synthesis flow. Our works have better timing results of up to 47% than synthesizing circuits with 0.13um static CMOS library developed by Faraday Corporation.

List of Contents:
Abstract 1
Contents 2
List of Figures 3
List of Tables 4
Chapter 1 Introduction 5
Chapter 2Partially Collapse for Timing Optimization 10
Chapter 3 Output Phase Assignment for Timing Optimization 16
Chapter 4 Remove Foot NMOS for Timing Optimization 21
Chapter 5 Pin Assignment for Timing Optimization 26
Chapter 6 Overall Flow 29
Chapter 7 Experimental Results 31
Chapter 8 Conclusions 34
References 35

List of Figures:
Fig. 1: A typical domino circuit 5
Fig. 2: Traditional domino synthesis flow 7
Fig. 3: The example of partially collapse for timing optimization 11
Fig. 4: The connection between each node and nodes in the t_slice list 13
Fig. 5: The final result of partially collapse nodes 14
Fig. 6: Different phase assignment for the same circuit 17
Fig. 7: Inverter trapped in the circuit 18
Fig. 8: Output phase assignment example 19
Fig. 9: The final result after performing output phase assignment 20
Fig. 10: 3-input domino AND cell 21
Fig. 11: DC current of the domino logic without foot NMOS 22
Fig. 12: The final result after performing output phase assignment 23
Fig. 13: The footless domino circuit with delay element 24
Fig. 14: Different pins have different speed 26
Fig. 15: The overall flow 29

List of Tables:
Table 1: The final experimental results 32

References
[1]. P. E. Gronowski, W. J. Bowhill, R. P. Preston, M. K. Gowan, and R. L. Allmon, “High-performance microprocessor design,” IEEE J. Solid-State Circuits, vol. 33, pp. 676–686, May 1998.
[2]. G. De Micheli, “Performance-oriented synthesis of large-scale domino CMOS circuits,” IEEE Trans. Computer-Aided Design, vol. CAD-6, pp. 751–765, Sept. 1987.
[3]. V. Friedman and S. Liu, “Dynamic logic CMOS circuits,” IEEE J. Solid-State Circuits, vol. SC-19, pp. 263–266, Apr. 1984.
[4]. R. Puri, A. Bjorksten, and T. Rosser, “Logic optimization by output phase assignment in dynamic logic synthesis,” in IEEE/ACM Int. Conf. Computer-Aided Design, 1996, pp. 2–8.
[5]. M. R. Prasad, D. Kirkpatrick, R. K. Brayton, and A. L. Sangiovanni-Vincentelli, “Domino logic synthesis and technology mapping,” Proc. Int. Workshop on Logic Synthesis, May 1997, vol. 1.
[6]. Ki-Wook Kim, Taewhan Kim, Liu, C.L., Sung-Mo Kang, “Domino logic synthesis based on implication graph,” Computer-Aided Design of Integrated Circuits and Systems, IEEE Transactions on Volume 21, Issue 2, Feb. 2002 Page(s):232 - 240
[7]. P. Patra and U. Narayanan, “Automated phase assignment for the synthesis of low power domino circuits,” in Proc. ACM/IEEE Design Automation Conf., 1999, pp. 379–384.
[8]. Thorp, T., Gin Yee, Sechen, C, “Domino logic synthesis using complex static gates,” Computer-Aided Design, 1998. ICCAD 98. Digest of Technical Papers. 1998 IEEE/ACM International Conference on 8-12 Nov 1998 Page(s):242 – 247.
[9]. Gin Yee, Sechen, C., “Dynamic logic synthesis,” Custom Integrated Circuits Conference, 1997., Proceedings of the IEEE 1997 5-8 May 1997 Page(s):345 – 348.
[10]. Gin Yee, Sechen C., “Clock-delayed domino for dynamic circuit design,” Very Large Scale Integration (VLSI) Systems, IEEE Transactions on Volume 8, Issue 4, Aug. 2000 Page(s):425 – 430.
[11]. G. Yee and C. Sechen, “Clock-delayed domino for adder and combinational logic design,” in Proc. IEEE/ACM Int. Conf. Computer Design, Oct. 1996, pp. 332–337.
[12]. Min Zhao, Sapatnekar, S.S. “Timing-driven partitioning and timing optimization of mixed static-domino implementations,” Computer-Aided Design of Integrated Circuits and Systems, IEEE Transactions on Volume 19, Issue 11, Nov. 2000 Page(s):1322 – 1336.
[13]. Puri, R., “Design issues in mixed static-domino circuit implementations,” Computer Design: VLSI in Computers and Processors, 1998. ICCD '98. Proceedings., International Conference on 5-7 Oct. 1998 Page(s):270 – 275.
[14]. E. Sentovich, K. Singh, L. Lavagno, C. Moon, R. Murgai, A. Saldanha, H. Savoj, P. Stephan, R. Brayton, and A. Smgiovaani-Viiimntelli, “SIS: a system for sequential circuit synthesis," Technical Report IJGB/ERL M92/41, University of Califoitnia, Berkeley, CA, May 1992.
[15]. John P. Uyemura, Introduction to VLSI circuits and systems, WILEY, 1st edition, 2002.
[16]. Bradley S. Carlson、C.Y. Roger Chen, “Performance enhancement of CMOS VLSI Circuits by transistor reordering,” in Proc. ACM/IEEE Design Automation Conf., 1993, pp. 361–366.