Method of determining delay in logic cell models

• US 6,028,995 A
• Filed: 03/31/1998
• Issued: 02/22/2000
• Est. Priority Date: 03/31/1998
• Status: Expired due to Term

• Current Assignee: Bell Semiconductor, LLC (Hilco Inc. (d/b/a Hilco Global))
• Original Assignee: LSI Logic Corporation (Broadcom, Inc.)
• Inventors: Jayasumana, Anura P., Jetton, Mark W.
• Primary Examiner(s): Teska, Kevin J.
• Assistant Examiner(s): Fiul, Dan
• Application Number:
  US09/052,914
• Time in Patent Office: 693 Days
• Field of Search: 395/500.07, 395/500.34, 395/500.35, 395/500.36, 395/500.4, 395/500.12, 395/500.44, 324/681
• US Class Current: 703/19
• CPC Class Codes: G06F 30/33 Design verification, e.g. functional simulation or model checking

Abstract

A logic-cell model accounts for nonlinear effects in determining propagation delay, thereby providing improved accuracy as compared to existing models, particularly when rise/fall times exceed several nanoseconds. Given a logic cell of the type wherein delay is a function of rise/fall time (TRL) and load capacitance (CL), the method involves choosing a plurality of discrete simulation points associated with the delay, each point also being a function of TRF and CL, after which the delay is determined in accordance with the chosen simulation points. One or more of the simulation points are preferably chosen in conjunction with both the linear and nonlinear regions of the TRL/CL space to ensure accuracy for a wide range of TRL and/or CL values. In the event of an identifiable or discontinuous transition between the linear and nonlinear regions, a discrete simulation point is also chosen with respect to the transition area. Based upon the simulation points, the invention is used to determine a plurality of constants which are then, in turn, used to solving for propagation delay on a more accurate basis. In this respect, the propagation delay, TD, may be determined in accordance with the relation

TD=A+B*CL+C*TRF+E(TRF,CL),

where A, B, C and E are constants from the simulation points, and with E(TRF,CL) representing a correction factor associated with the nonlinear effects.

Claims

1. A method of simulating the timing of an integrated circuit by predicting the propagation delay of a signal through a logic cell of the type wherein the delay is a function of a rise/fall time TRF and load capacitance (CL), the method comprising the steps of:

choosing a plurality of discrete simulation points defined at least in part by particular values of TRF and CL;

determining the delay for each of the chosen simulation points, thereby generating a plurality of delay values;

using the plurality of delay values to predict the propagation delay for values of TRF and CL other than the particular values of TRF and CL associated with the plurality of discrete simulation points; and

using the predicted propagation delays to simulate the timing of an integrated circuit.

2. The method of claim 1, wherein the delay as a function of TRF and CL exhibits a transition between a linear region and a nonlinear region, and wherein the step of choosing a plurality of discrete simulation points associated with the delay includes choosing one or more points associated with the linear region and one or more points associated with the nonlinear region.

3. The method of claim 1, wherein the delay as a function of TRF and CL exhibits a transition between a linear region and a nonlinear region, and wherein the step of choosing a plurality of discrete simulation points associated with the delay includes choosing one or more points associated with the transition between the linear and nonlinear regions.

4. The method of claim 1, wherein the logic cell is a single-stage cell.

5. The method of claim 4, wherein the step of choosing a plurality of discrete simulation points includes choosing the following points:

a) a first simulation point corresponding to values of TRF within a first range and CL value within a second range;

b) a plurality of different simulation points, each corresponding to a TRF value within the first range and a value of CL within a third range, each of the values in the third range being higher than any value in the second range; and

c) a plurality of different simulation points, each corresponding to a TRF value within a fourth range and a value of CL within the second range, each of the values in the fourth range being higher than any value in the first range.

6. The method of claim 1, wherein the logic cell is a multi-stage cell.

7. The method of claim 1, further including the step of determining a plurality of constants in accordance with the simulation points, and wherein the step of solving for the propagation delay incorporates the constants.

8. A method of simulating the timing of an integrated circuit by predicting the propagation delay of a signal through a logic cell which accounts for both linear and nonlinear effects, the cell being of the type wherein the delay is a function of a rise/fall time TRF and load capacitance (CL), the method comprising the steps of:

choosing a plurality of discrete simulation points defined at least in part by particular values of TRF and CL;

determining the delay TD for each of the chosen simulation points, thereby generating a plurality of delay values;

solving for constants in an equation based on the plurality of discrete simulation points and associated plurality of delay values, wherein the equation is

space="preserve" listing-type="equation">TD=A+B*CL+C*TRF+E(TRF,CL),

where A, B, C and E are the constants based on the plurality of discrete simulation points and associated plurality of delay values, and

E(TRF,CL) is a correction factor associated with the nonlinear effects; and

using the equation to simulate the timing of an integrated circuit.

9. The method of claim 8, wherein, for single-stage cells:

##EQU4##10.

10. The method of claim 8, wherein, for multi-stage cells:

space="preserve" listing-type="equation">E(TRF,CL)=D*(TRF)1/2+(E*TRF+F*(TRF)1/2)/(Max(CL,cl))1/2.