Use

To improve system performance during the planning run, you can specify that parallel processing is to be used when planning.

For products with many component variants, you can also specify that the system is to create several sub-matrices for a product instead of one very large matrix. Splitting one large matrix into several sub-matrices accelerates the planning run only when done in conjunction with parallel processing.

Integration

You make the settings for parallel processing in the global settings , or on the initial screen for the planning run for the planning matrix.

Prerequisites

In Customizing for Production Planning and Detailed Scheduling (PP/DS), in the IMG activity Maintain Destinations for Parallel Processing, you have entered the server you want to use for planning .

If you want to split large matrices into several sub-matrices, you must specify how many sub-matrices are required in the Control Parameters for RPM Products . You must do so for each product that is to be planned using the planning matrix.

Features

Parallel processing

During the planning run, the planning matrix should always be created using parallel processing.

To ensure this, go to the Global Settings and in the Parallel Processing field, choose RPM part of planning run is always processed in parallel. If you have selected this setting in Customizing, the part of the planning run in which the planning matrix is created for a product is always executed using parallel processing, even if you have not set the indicator for parallel processing on the initial screen for the planning run.

Parallel processing with creation of sub-matrices

Sub-matrices should be created only if the planning run is executed using parallel processing.

To ensure this, go to the Global Settings and in the Parallel Processing field, choose Sub-matrices only in parallel processing . Set the indicator for parallel processing on the initial screen for the planning matrix run. See also: Updating Matrices .

We recommend you to use sub-matrices if the product requires a large number of components. The planning matrix is then split into several matrices. To improve performance when the matrix is split into several matrices, parallel processing should be used for the planning run.