Deep Submicron Interconnect Timing Model with Quadratic Random Variable Analysis

Zeng, Jun-Kuei; Chen, Chung-Ping

doi:10.1109/date.2008.4484922

Cited by 3 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The quadratic timing model has a good accuracy as compared to Monte Carlo results. More accurate than the result in [6]. And our SMC used in the Statistical timing analysis greatly speed up the time of getting the moments and its derivatives.…”

Section: Application To Statistical Timing Analysismentioning

confidence: 95%

See 1 more Smart Citation

Symbolic moment computation with application to statistical timing analysis

Xie

2009

2009 Asia Pacific Conference on Postgraduate Research in Microelectronics &Amp; Electronics (PrimeAsia)

View full text Add to dashboard Cite

where General RLC circuits can be modeled in the following form where the state vector x(t) comprises of the node voltages and branch currents necessary for circuit analysis. If the circuit is of the regular tree structure as shown in Fig 1, then the system matrices E, A and F in equation 1 would be of special structure, which simplifies the moment computation .Circuit moment is a concept derived from the expansion of the frequency response H (s) of a linear network in a series of s, for example in Section III. Application of SMC in statistical timing analysis is shown in Section IV. Conclusions are made in Section V is called the ith moment of the impulse response h(t) for i = 0,1 ,2, ' . '. Here for convenience we have defined all moments to be nonnegative. Moments are useful in that they can be used for constructing reduced order models with the constraint of matching low-order moments [1] [7].

show abstract

Section: Application To Statistical Timing Analysismentioning

confidence: 95%

“…(BDD Evaluation Phase) The value at each BDD root node R, is equal to the algebraic value defined recursively by the following formula: (6) where val (Zero) = 0 and val(R i --+ Then) = C(i). After the recursion, val(R ) gives rise to the moment m(i ,j).…”

Section: Smc Moment Evaluation Algorithmmentioning

confidence: 99%

Symbolic moment computation with application to statistical timing analysis

Xie

2009

2009 Asia Pacific Conference on Postgraduate Research in Microelectronics &Amp; Electronics (PrimeAsia)

View full text Add to dashboard Cite

show abstract

“…Along this line, recently we have observed that quite a number of researches have attempted to extend the model order reduction techniques to considering the variation issues. Typical approaches, among others, are 1) Incorporating interval analysis to matrix computation, resulting interval algebra for reduced-order modeling [3]; 2) Symbolic model order reduction techniques [4] which are far from mature; and 3) Other approximate or parametric approaches formulated under the name of statistical timing analysis, such as [5,6]. Obviously, the model order reduction based approaches are confronting lots of difficulties as long as mesh networks are concerned.…”

Section: Introductionmentioning

confidence: 99%

A fast symbolic computation approach to statistical analysis of mesh networks with multiple sources

Hao¹,

Shi²

2010

2010 15th Asia and South Pacific Design Automation Conference (ASP-DAC)

View full text Add to dashboard Cite

Mesh circuits typically consist of many resistive links and many sources. Accurate analysis of massive mesh networks is demanding in the current integrated circuit design practice, yet their computation confronts numerous challenges. When variation is considered, mesh analysis becomes a much harder task. This paper proposes a symbolic computation technique that can be applied to the moment-based analysis of mesh networks with multiple sources. The variation issues are easily taken care of by a structured computation mechanism, which can naturally facilitate sensitivity based analysis. Applications are addressed by applying the computation technique to a set of mesh circuits with varying sizes.

show abstract