考慮存在電路內部的耦合電容找到最大電流值

Finding Maximum Currents Considering Interconnect Coupling Capacitance

何京霖(Ching-Lin Ho)

指導教授 ： 林榮彬

元智大學/資訊學院/資訊工程學系/碩士(2004年)

摘要

在超大型積體電路設計的領域裡，如何在電源分布網路中精確地測量到最大電流值是非常必要的。瞬間的電流極值支配著瞬間電壓降的程度，過度的電壓降會造成晶片無法達到正常功能的運作。因此，能夠量得精確的最大電流值是很重要的。在本篇論文中，我們將提出如何找到最大電流值的方法，進一步地去解決過度壓降的問題。由於製程技術的提昇，不斷地深入深次微米的層次，兩條導線間的耦合電容變成了影響電流的一個重要因素。過大的耦合電容不僅僅會引起雜訊干擾，在訊號延遲上的程度也會增加，這會使得電流量測方面的困難度提高許多。於是，我們的方法也同時考慮了耦合電容這一重要影響因素。首先，我們會在每條導線上產生一近似的電流波形，再利用重疊法將各個電流波形相加，並找出在合成波形中的最大電流值。實驗數據顯示，在單一aggressor的影響下，我們的方法可以產生精確的最電大流值，平均誤差大概在12.5%。但是，當aggressor數目超過兩條以上時，誤差值卻到達29.3%，比在單一aggressor的情況下大了一倍以上。未來的重要課題便是如何去提升在多個aggressor影響下的效能且能被應用在較大規模的電路設計上。

關鍵字

耦合電容 ； 最大電流值

並列摘要

Accurate maximum current estimation in a VLSI design is essential to a successful power network design. The peak current in power supply network dictates the degree of voltage drop on power supply. Excessive voltage drop on the power supply for a VLSI design could lead to chip failure. Therefore, it is very important to have accurate peak current estimation. Current estimation is not an easy task because the current magnitude is highly dependent on the input pattern, gate type, slew rate of a signal, arrival time of signal, etc. In this thesis, we propose a methodology to perform accurate estimation of the maximum instantaneous current. As VLSI technology continues to push toward deep submicron, the coupling capacitance between adjacent wires has become the dominating component for current estimation. High coupling capacitance in deep submicron design results in noise, additional delay, and makes the current and timing estimation is increasingly intractable. It is utmost important to include coupling capacitance along with self-capacitance for performing a more accurate maximum current estimation. In our approach, we find out the related current waveform for a net at first, and calculate the maximum total current value by superimposition method. We use curve-fitting method to obtain some closed-form equations which can be used to reconstruct the related current waveform for fast maximum current estimation. Finally, we utilize a circuit with two inverter-chains to perform our experiments. The experimental results show that our approach could derive a good approximate value for maximum current in single-aggressor model. The average errors are about with respect to the current value obtained by HSPICE. However, there are about 29.3% of average errors for multi-aggressor model. Future work is proposed to improve the accuracy for multi-aggressor model.

並列關鍵字

coupling capacitance ； Maximum current

參考文獻

[Krip95] H. Kriplani, F. N. Najm and I. N. Hajj, “Pattern Independent Maximum Current Estimation in Power and Ground Buses of CMOS VLSI Circuits：Algorithm, Signal Correlations, and Their Resolution”, Computer-Aided Design of Integrated Circuits and Systems, IEEE Transactions on, Volume: 14, Issue: 8, Aug. 1995. Pages: 998 — 1012.

[Krst97] A. Krstic and K. T. Cheng, “Vector Generation for Maximum Instantaneous Current Through Supply Lines for CMOS Circuits”, Design Automation Conference, 1997. Proceedings of the 34th, June 9-13, 1997. Pages: 383 — 388.

[Jia97] Y. M. Jiang and K. T. Cheng, “Estimation of Maximum Power and Instantaneous Current Using a Genetic Algorithm”, in Proc. of IEEE Custom Integrated Circuits Conference. May 1997. Pages: 135 — 138.

[Dev90] S. Devadas, K. Keutzer and J. White, “Estimation of Power Dissipation in CMOS Combinational Circuits”, Custom Integrated Circuits Conference, Proceedings of the IEEE 1990, 13-16 May 1990. Pages: 19.7/1 - 19.7/6.

[Jia98] Yi-Min Jiang and Kwang-Ting Cheng, “Exact and Approximate Estimation for Maximum Instantaneous Current of CMOS Circuits”, Design Automation and Test in Europe, Proceedings, 23-26 Feb. 1998. Pages: 698 — 702.