摘要:
A method, system, and computer program product are disclosed for performing statistical leakage power characterization to estimate yield of a circuit in terms of leakage power. According to some approaches, this is performed with consideration of state correlation.
摘要:
A method, system, and computer program product are disclosed for performing statistical leakage power characterization to estimate yield of a circuit in terms of leakage power. According to some approaches, this is performed with consideration of state correlation.
摘要:
In one embodiment of the invention, a method of analysis of a circuit design with respect to within-die process variation is disclosed to generate a design-specific on chip variation (DS-OCV) de-rating factor. The method includes executing a static timing analysis (STA) in an on-chip variation mode using a process corner library. Collecting timing information of the top N critical timing paths. Executing a statistical static timing analysis (SSTA) on the N critical timing paths using timing models characterized for SSTA with sensitivities of delays to process variables. Compare the two timing results and deriving DS-OCV de-rating factors for the clock/data paths to be used in a STA OCV timing analysis to correctly account for the effects of process variations. A user may select to specify DS-OCV de-rating factors for paths or groups of paths and achieve an accurate timing analysis report in a reduced amount of run-time.
摘要:
In one embodiment of the invention, a method of analysis of a circuit design with respect to within-die process variation is disclosed to generate a design-specific on chip variation (DS-OCV) de-rating factor. The method includes executing a static timing analysis (STA) in an on-chip variation mode using a process corner library. Collecting timing information of the top N critical timing paths. Executing a statistical static timing analysis (SSTA) on the N critical timing paths using timing models characterized for SSTA with sensitivities of delays to process variables. Compare the two timing results and deriving DS-OCV de-rating factors for the clock/data paths to be used in a STA OCV timing analysis to correctly account for the effects of process variations. A user may select to specify DS-OCV de-rating factors for paths or groups of paths and achieve an accurate timing analysis report in a reduced amount of run-time.
摘要:
A method, system, and computer program product are disclosed for performing crosstalk analysis using first-order parameterized analysis modeling. The approach can be used to factor in the effect of process variations within the definition of timing windows. This approach allows one to bypass the simplistic assumptions related to best-case/worst-case analysis using timing windows, and provide a realistic picture of the impact of timing windows on noise analysis. The timing windows can be viewed in terms of the individual process parameter. The process parameters could be real process parameters, or virtual/computed components based on the actual process parameters. The process parameters can be used to compute overlap of timing windows for performing noise analysis.
摘要:
Techniques are presented for determining effects of process variations on the leakage of an integrated circuit having multiple devices. The operation of the circuit is simulated using a first set of values for the process parameters for the devices and is also simulated with some of the process parameter values varied. For the simulation with the varied values, the circuit is split up into distinct components (such as channeled coupled components, CCCs), where each component has one or more devices, and a process parameters value in a device in each of two or more of these components is varied.
摘要:
Techniques are presented for determining effects of process variations on the leakage of an integrated circuit having multiple devices. The operation of the circuit is simulated using a first set of values for the process parameters for the devices and is also simulated with some of the process parameter values varied. For the simulation with the varied values, the circuit is split up into distinct components (such as channeled coupled components, CCCs), where each component has one or more devices, and a process parameters value in a device in each of two or more of these components is varied.
摘要:
In one embodiment of the invention, a method for electronic circuit design is disclosed. The method includes analyzing a netlist of a subcircuit to determine one or more input pins and one or more output pins; forming an arc graph of the subcircuit including one or more timing arcs between the one or more input pins and the one or more output pins; and reducing the number of transistors to perturb to perform a sensitivity analysis for within die process variations over the one or more timing arcs to reduce the number of simulations to characterize the subcircuit.
摘要:
A method, system, and computer program product are disclosed for performing crosstalk analysis using first-order parameterized analysis modeling. The approach can be used to factor in the effect of process variations within the definition of timing windows. This approach allows one to bypass the simplistic assumptions related to best-case/worst-case analysis using timing windows, and provide a realistic picture of the impact of timing windows on noise analysis. The timing windows can be viewed in terms of the individual process parameter. The process parameters could be real process parameters, or virtual/computed components based on the actual process parameters. The process parameters can be used to compute overlap of timing windows for performing noise analysis.