面试系列:可测性设计【3】关于测试覆盖率

  Why Design-for-Test?

       To increase Productivity:

            Shorter time-to-market

            Reduced design cycle

            Reduced cost

       To improve Quality:

            Reduced Defects per million(DPM)

            Improved quality of test

  What is Testability?

        Controllability:

            The ability to set a node to aspecific value

        Observability:

            The ability to observe a node’s value

Fault Models

         Fault models:

             Stuck-at-fault

             Transition fault

             Path delay

             IDDQ

  Testable fault classes

   Detected (DT)

               det_simulation (DS)

                    faults detected when the toolperforms fault simulation

               det_implication (DI)

                    faults detected when the toolperforms learning analysis

           The list are only for pathdelay testing

                det_robust (DR)

                      robust detected faults

                det_functional (DF)

                     functionally detected faults

  Testable fault classes

          Posdet (PD):

           It includes all faults thatfault simulation identifies as possible-detected but not

            hard detected

           The posdet class contains twogroups:

                 posdet_testable (PT) -potentially detectable posdet faults. A higher abort

                  limit may reduce thenumber of these faults

                 posdet_untestable (PU) - provenAU during pattern generation and hard

                  undetectable posdet faults

            By default, the calculationsgive 50% credit for posdet faults