摘要:
The present invention is a system and method for efficiently and incrementally updating the statistical timing of a digital circuit after a change has been made in the circuit. One or more changes in the circuit is/are followed by timing queries that are answered efficiently, constituting a mode of timing that is most useful in the inner loop of an automatic computer-aided design (CAD) synthesis or optimization tool. In the statistical re-timing, the delay of each gate or wire is assumed to consist of a nominal portion, a correlated random portion that is parameterized by each of the sources of variation and an independent random portion. Correlations are taken into account. Both early mode and late mode timing are included; both combinational and sequential circuits are handled; static CMOS as well as dynamic logic families are accommodated.
摘要:
The present invention is a system and method for determining criticality probability of each node, edge and path of the timing graph of a digital circuit in the presence of delay variations. The delay of each gate or wire is assumed to consist of a nominal portion, a correlated random portion that is parameterized by each of the sources of variation and an independent random portion. Correlations are taken into account. Both early mode and late mode timing are included; both combinational and sequential circuits are handled; static CMOS as well as dynamic logic families are accommodated. The criticality determination complexity is linear in the size of the graph and the number of sources of variation. The invention includes a method for efficiently enumerating the critical path(s) that is/are most likely to be critical.
摘要:
Disclosed is a method for achieving timing closure in the design of a digital integrated circuit or system by selecting portions of the circuit or system to be optimized and portions of the circuit or system in which the effects of such optimization are to be analyzed during the optimization process. Optimized portions will include gates whose design parameters are to be changed, a first analyzed portion includes gates whose delays and edge slews are to be recomputed, and a second analyzed portion includes gates whose ATs and RATs are to be recomputed during optimization. Constraints are imposed at selected boundaries between these portions to prevent unwanted propagation of timing information and to ensure the validity of timing values used during optimization. Through this selection, the size of the problem posed to the underlying optimization method will be reduced, allowing larger circuits or systems to be optimized and allowing optimization to be performed more quickly.
摘要:
Solutions for ordering of statistical correlated quantities are disclosed. In one aspect, a method includes timing a plurality of paths in an integrated circuit to determine a set of timing quantities associated with each of the plurality of paths; determining a most critical timing quantity in the set of timing quantities; forming a tiered timing quantity arrangement for ordering a plurality of timing quantities in the set of timing quantities; removing the most critical timing quantity from the set of timing quantities and placing the most critical timing quantity in an uppermost available tier of the tiered timing quantity arrangement; and repeating the determining, forming and removing for the set of timing quantities excluding the removed most critical timing quantity.
摘要:
Disclosed is a method for enhanced efficiency and effectiveness in achieving closure of large, complex, high-performance digital integrated circuits. Circuit macros are re-optimized and re-tuned in the timing closure loop by means of a reformulated objective function that allows the optimizer to improve the slack of all signals rather than just the most critical one(s). The incentive to improve the timing of a sub-critical signal is a diminishing function of the criticality of the signal. Thus all signals are improved during the optimization, with the highest incentive to improve on the most critical signals, leading to faster and more effective overall timing closure.
摘要:
A method, system, and integrated circuit including selectively added timing margin. The method, for integrating statistical timing and automatic test pattern generation (ATPG) to selectively add timing margin in an integrated circuit, includes identifying, while a chip is in design, paths that are unable to be robustly tested “at speed” during manufacturing test, running statistical timing to calculate a margin to be applied to the paths, updating design specifications for margin to be applied to the paths, and optimizing chip logic based on updated design specifications.
摘要:
The present invention is a system and method for efficiently and incrementally updating the statistical timing of a digital circuit after a change has been made in the circuit. One or more changes in the circuit is/are followed by timing queries that are answered efficiently, constituting a mode of timing that is most useful in the inner loop of an automatic computer-aided design (CAD) synthesis or optimization tool. In the statistical re-timing, the delay of each gate or wire is assumed to consist of a nominal portion, a correlated random portion that is parameterized by each of the sources of variation and an independent random portion. Correlations are taken into account. Both early mode and late mode timing are included; both combinational and sequential circuits are handled; static CMOS as well as dynamic logic families are accommodated.
摘要:
In one embodiment, the invention is a method and apparatus for static timing analysis in the presence of a coupling event and process variation. One embodiment of a method for computing a statistical change in delay and slew due to a coupling event between two adjacent nets in an integrated circuit design includes conducting a statistical timing analysis of the integrated circuit design, computing a statistical overlap window between the adjacent nets, where the statistical timing window represents a period of time during which signals on the adjacent nets can switch contemporaneously and computing the statistical change of delay due to the coupling event, in accordance with the statistical overlap window.
摘要:
A computer system for partitioning the columns of a matrix A. The computer system includes a processor and a memory unit coupled to the processor. Program code in the memory unit, when executed by the processor, implements the method. Matrix A is provided in a memory device and has n columns and m rows; wherein n is an integer of at least 3; and wherein m is an integer of at least 1. The n columns is partitioned into a closed group of p clusters, p being a positive integer of at least 2 and less than n. The partitioning includes an affinity-based merging of clusters of pairs of clusters of the matrix A based on an affinity between the clusters in each pair of clusters being merged. Each cluster consists of one or more columns of matrix A. The p clusters are stored in a computer-readable storage device.
摘要:
A method for partitioning the columns of a matrix A. The method includes providing the matrix A in a memory device of a computer system. The matrix A has n columns and m rows, wherein n is an integer of at least 3, and wherein m is an integer of at least 1. The method further includes executing an algorithm by a processor of the computer system. Executing the algorithm includes partitioning the n columns of the matrix A into a closed group of p clusters, wherein p is a positive integer of at least 2 and less than n, wherein the partitioning includes an affinity-based merging of clusters of the matrix A, and wherein each cluster is a collection of one or more columns of A.