Run Evaluation

Select which runs to evaluate and which tasks to perform.

1. Go to the Evaluate step.
2. In the Evaluation Tasks tab, Active column, select the runs to evaluate.
3. In the Run Tasks tab, select the checkboxes of the tasks to perform.
   By default, Write Input Files, Execute Analysis, and Extract Output Responses are active.

   
   
   Figure 1.

4. Define optional settings.

   Setting	Action
   Notification of task completion	Click and activate Notify.
   Write solver output in Message Log and/or log-file	Click and activate Log External Output.
   Change the number of concurrent jobs to run	Click Multi-Execution and enter a new value; doesn't have to be a static entry. Enter 0 to stop the submission of new jobs. Click to select an execution mode.

   Multi-execute is a job management setting used to control throughput. Some algortithm's specfication settings can affect the number of jobs created per iteration. To ensure repeatability, the two settings are not tied together. However, it is recommended to coordinate the settings to ensure maximum use of resources.

   A Fit can run in multi-execute while simultaneously iterating over multiple responses.

   Multi-execution runs jobs in vertical, horizontal, or horizontal (write all first) execution mode.

   • Vertical execution mode performs the write, execute, and extract tasks for all designs simultaneously; that is all designs are written, then executed, then extracted.
   • Horizontal execution mode sequences the write, execute, and extract task for each run independently.
   • Horizontal (write all first) execution mode sequences the write task for each run first, then sequences the execute and extract tasks for each run independently.
5. Click Evaluate Tasks.