Method of identifying an accurate model
AssigneeSTMicroelectronics S.r.l.
InventorPeter Ballam
Life of this patent
Estimating the term…
Abstract
A method is described for identifying an inaccurate model of a hardware circuit. The method includes the steps of simulating the model of the circuit by applying a plurality of signals, said plurality of signals having at least one abstract data type level to provide a set of expected results, replacing the at least one abstract data type level with two or more levels having different values to thereby provide and expanded set of signals to apply to said model, resimulating the model with said expanded set and comparing the two sets of results and providing an output signal indicating if the model is inaccurate if the results contradict.
Claims12
A method of identifying an inaccurate model of a hardware circuit comprising the steps of:
- running the model of the circuit by applying a plurality of signals to the model to obtain a first set of expected results, said plurality of signals having at least one abstract data type level;
- replacing the at least one abstract data type level with two or more levels having different values to thereby provide an expanded set of signals to apply to said model;
- re-running the model by applying said expanded set of signals to the model to obtain a second set of results;
- and
- comparing the first set of expected results with the second set of results and providing an output signal indicating that the model is inaccurate if the first set of expected results and the second set of results contradict.
A system for identifying an inaccurate model of a hardware circuit comprising:
- means for running the model of the circuit by applying a plurality of signals to the model to obtain a first set of expected results, said plurality of signals having at least one abstract data type level;
- means for replacing the at least one abstract data type level with two or more levels having different values to thereby provide an expanded set of signals to apply to said model;
- means for re-running model by applying said expanded set of signal to the model to obtain a second set of results;
- and
- means for comparing the first set of expected results with the second set of results and providing and output signal indicating that the model is inaccurate if the first set of expected results and the second set of results contradict.
Dependencies are inferred from the claim text.