Optimization Tab

Lets you specify the optimization algorithm, starting point of the simulation run, evaluation type, and stopping criteria.

Monte Carlo Tab

Lets you specify the Monte Carlo run options for the Improve Yield mode.

Algorithm

Selects one of the following optimization algorithms:

• Global Optimization
• BFGS
• Conjugate Gradient
• Brent-Powell
• Hooke-Jeeves

Starting Point

Selects one of the following options to specify a starting point for the simulation run:

• Use Starting State: To use the setup state as the starting point for the optimization. Select a setup state that defines a set of fixed values for every global variable or parameter that defines a range of values in the active setup.
  You must have a setup state available to use this option.
  
  If the setup of your cellview earlier used a reference point as a starting point, the details of the starting point have been saved in a setup state named ref_point_migrated. Name of this state is automatically selected in the Use Starting State drop-down list on the form.
• No (Runs Initial Global Optimization): To perform an initial global optimization on the nominal corner and to use its best point as the starting point for the run. If you select this option, ensure that:
  • The nominal corner is not disabled in the Run Summary pane.
  • For tests that are enabled in the Data View pane, the nominal corner is not disabled in the Corners Setup form.

Evaluation

Lets you select one of the following evaluation methods:

• Full
• Conditional

For more information, see Full Evaluation and Conditional Evaluation.

Stopping Criteria

Lets you use either the recommended options or custom options for stopping criteria.

Time Limit (minutes)

Sets a tile limit for the run.

Number of Iterations

Specifies the number of sizing/Monte Carlo iterations.

The default is 3.

Max Points Per Iteration

Specifies the maximum number of points processed per iteration.

The default is 3000.

Stop Iteration Early If No Improvement

Stops the process early if the sizing results in no improvement.

This option is applied to each optimization iteration. This stopping criteria is similar to the No Improvement with Points stopping criteria for the global or local optimization run modes for which the user also specifies the number of points. In the case of iterative run modes, the number of points is calculated as Max Points Per Iteration / 3.

Statistical Variation

Selects one of the following statistical variation options:

• Mismatch: For per-instance statistical variations.
• Process: For process statistical variations.
• All: For both process and per-instance statistical variations.

Sampling Method

Selects one of the following statistical sampling methods supported by Spectre:

• Random
• Latin Hypercube
• Low-Discrepancy Sequence

For more information, see Statistical Sampling Methods.

Number of Points

Specifies the number of Monte Carlo points to be simulate.

Number of Bins

Specifies the number of bins (subdivisions) for the Latin Hypercube method.

• If a number is specified, the number of bins will be the specified number, or (Number of Points + Starting Run Number - 1), whichever is greater.
  For example, if the specified number of bins is 90, the number of points specified in the Number of Points field is 100 and the starting run number specified in the Starting Run Number field is 6, the value 105 (100+6-1) is used.
• If no number is specified, a default value of (Number of Points + Starting Run Number - 1) is used.
  For example, if the number of points specified in the Number of Points field is 100 and the starting run number specified in the Starting Run Number field is 6, the default value of 105 (100+6-1) is used.

Run Normal Simulation

Runs normal simulation.

Monte Carlo Seed

Specifies a seed for the Monte Carlo analysis.

By always specifying the same seed, you can reproduce a previous experiment.

If you do not specify a seed, the value 12345 is used by default.

Starting Run Number

Specifies the run that Monte Carlo begins with.

By specifying this number, you can reproduce a particular run or sequence of runs from a previous experiment, for example, to examine an earlier case in detail.

Specify Instances/Devices

Click to specify the sensitive instances and devices you want to either include or exclude for applying mismatch variations.

For more information, see Specifying Instances for Monte Carlo Mismatch and Process Variation in Virtuoso ADE Explorer User Guide.

Specify Tests for Monte Carlo

Click to select the tests for which you want to run Monte Carlo simulation.