June 7, 2016: Large, complex models can be distributed across multiple processing units and processor cores to ensure the simulation runs in real time. There are two different workflows for this. The first is to use separate Simulink® models for each core and import them into ConfigurationDesk. In this workflow, intermodel communication is configured in ConfigurationDesk.
In the second workflow, there is one overall Simulink® model for the whole application, and a special Simulink block is used to specify which of its subsystems should be computed together on one core. The overall model is then automatically split into separate model files. In this workflow, intermodel communication is transferred from Simulink® to ConfigurationDesk.
One processor core executes one model. Several models are combined into processing-unit applications. These can be assigned to processing units in ConfigurationDesk, which automatically performs the core-to-model assignment within each unit.
One processing unit consists of several processor cores. One processor core in each processing unit is always reserved for communication with the host PC. The other cores can be used to calculate the behavior model.