Variable accuracy parameter modeling in statistical timing
Systems and methods compute a mean timing value of an integrated circuit design for variables using a first timing calculation of relatively higher accuracy; and calculate a first timing value of the integrated circuit design for the variables, using a second timing calculation having a relatively lower accuracy. Such systems and methods calculate second timing values of the integrated circuit design for additional sets of variables using the second timing calculation; and calculate finite differences of each of the second timing values to the first timing value. Thus, these systems and methods calculate a statistical sensitivity of the first timing value to the additional sets of variables based on the finite differences. Further, such systems and methods calculate a statistical sensitivity of the mean timing value to the additional sets of values based on the statistical sensitivity of the first timing value to the additional sets of values.
1. A method for performing statistical static timing analysis for an integrated circuit design comprising:
receiving said integrated circuit design into a computer;
automatically computing at least one mean timing value of said integrated circuit design using first manufacturing process variables and a first timing calculation, using said computer;
automatically computing second timing values of said integrated circuit design using a second sets of manufacturing process variables and a second timing calculation, using said computer, said second timing calculation having a lower accuracy relative to said first timing calculation;
automatically calculating finite differences of each of said second timing values to said at least one mean timing value, using said computer; and
automatically computing at least one statistical sensitivity of said at least one mean timing value to said second sets of manufacturing process variables based on said finite differences, using said computer to improve operating speed and decrease malfunction of integrated circuits fabricated according to said integrated circuit design.
2. The method according to claim 1 , said first timing calculation comprising at least one of:
current source modeling,
Simulation Program with Integrated Circuit Emphasis (SPICE) modeling,
input pin capacitance (pincap) modeling as a function of manufacturing process variations, and
full effective capacitance (Ceff) calculation.
3. The method according to claim 1 , said second timing calculations comprising at least one of:
tabular modeling;
base corner capacitance used for perturbed corners;
base corner effective capacitance used for perturbed corners; and
lumped capacitance modeling.
4. The method according to claim 1 , further comprising performing higher accuracy modeling for selected sensitivities, based on a pre-determined accuracy impact.
5. The method according to claim 1 , further comprising performing higher accuracy modeling for:
physical elements of said integrated circuit design subject to relatively increased accuracy requirements; and
hierarchical elements of said integrated circuit design subject to relatively increased accuracy requirements.
6. The method according to claim 1 , said first timing calculation comprising sensitivities, cross-terms, and opposite corner values.
7. A method comprising:
receiving an integrated circuit design into a computer;
automatically computing a mean timing value of at least one portion of said integrated circuit design for a first set of manufacturing process variables using a first timing calculation having a first accuracy and a first computer resource cost, using said computer;
automatically calculating second timing values of said at least one portion of said integrated circuit design for additional sets of manufacturing process variables that are different from said first set of manufacturing process variables using a second timing calculation having a second accuracy and having a second computer resource cost, using said computer;
automatically calculating finite differences of each of said second timing values to said mean timing value, using said computer;
automatically calculating a statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables based on said finite differences, using said computer; and
automatically outputting said mean timing value and said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables from said computer for use in modifying said integrated circuit design to improve operating speed and decrease malfunction of integrated circuits fabricated according to said integrated circuit design,
said second accuracy being lower than said first accuracy, and
said second computer resource cost being lower than said first computer resource cost.
8. The method according to claim 7 , said automatically calculating second timing values further comprising using said first timing calculation for selected sets of said additional sets of manufacturing process variables, based on said selected sets of said additional sets of manufacturing process variables impacting accuracy of said statistical sensitivity of said mean timing value.
9. The method according to claim 7 , further comprising calculating said second timing values using said first timing calculation for selected portions of said integrated circuit design based on:
physical elements of said integrated circuit design subject to relatively increased accuracy requirements; and
hierarchical elements of said integrated circuit design subject to relatively increased accuracy requirements.
10. The method according to claim 7 , said automatically calculating a statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables comprising:
determining a distribution of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables; and
centering said distribution of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables on said mean timing value.
11. The method according to claim 7 , said outputting of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables comprising outputting a graph showing a distribution of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables centered on said mean timing value.
12. The method according to claim 7 , said first computer resource cost and said second computer resource cost comprising turn-around-time (TAT).
13. The method according to claim 7 , said first timing calculation and said second timing calculation comprising different circuit timing models.
14. A computer program product for performing statistical static timing analysis, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions being readable and executable by a computer to cause the computer to perform a method comprising:
receiving an integrated circuit design into a computer;
automatically computing a mean timing value of at least one portion of said integrated circuit design for a first set of manufacturing process variables using a first timing calculation having a first accuracy and a first computer resource cost, using said computer;
automatically calculating second timing values of said at least one portion of said integrated circuit design for additional sets of manufacturing process variables that are different from said first set of manufacturing process variables using a second timing calculation having a second accuracy and having a second computer resource cost, using said computer;
automatically calculating finite differences of each of said second timing values to said mean timing value, using said computer;
automatically calculating a statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables based on said finite differences, using said computer; and
automatically outputting said mean timing value and said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables from said computer for use in modifying said integrated circuit design to improve operating speed and decrease malfunction of integrated circuits fabricated according to said integrated circuit design,
said second accuracy being lower than said first accuracy, and
said second computer resource cost being lower than said first computer resource cost.
15. The computer program product according to claim 14 , said automatically calculating second timing values further comprising using said first timing calculation for selected sets of said additional sets of manufacturing process variables, based on said selected sets of said additional sets of manufacturing process variables impacting accuracy of said statistical sensitivity of said mean timing value.
16. The computer program product according to claim 14 , said method further comprising calculating said second timing values using said first timing calculation for selected portions of said integrated circuit design based on:
physical elements of said integrated circuit design subject to relatively increased accuracy requirements; and
hierarchical elements of said integrated circuit design subject to relatively increased accuracy requirements.
17. The computer program product according to claim 14 , said automatically calculating a statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables comprising:
determining a distribution of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables; and
centering said distribution of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables on said mean timing value.
18. The computer program product according to claim 14 , said outputting of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables comprising outputting a graph showing a distribution of said statistical sensitivity of said mean timing value to said additional sets of manufacturing process variables centered on said mean timing value.
19. The computer program product according to claim 14 , said first computer resource cost and said second computer resource cost comprising turn-around-time (TAT).
20. The computer program product according to claim 14 , said first timing calculation and said second timing calculation comprising different circuit timing models.