Performance Optimization
To improve the performance of your real-time application, consider the following changes:
Change a single-rate model that is inherently multirate to a multirate model.
Use a multicore target computer, and then set the tasking mode to multitasking.
Detect potential improvements for your model by using Performance Advisor.
Determine the lower limit on sample time that does not produce a CPU overload.
Allow a limited number of CPU overloads with the
TLCOptions Properties
.Configure a real-time application for profiling, run it, and call
profile_slrt
to retrieve the results. Evaluate the results for potential improvements in the task and core distribution of the model.For an inherently single-rate model, execute the real-time application on a multicore target computer with polling mode.
Distribute processing among target computers connected by a network.
Delegate high sample rate I/O to an FPGA subsystem.
- Profiling and Optimization
Execution profiling, display, analysis, and optimization by using Performance Advisor and MATLAB®功能 - Concurrent Execution with Multicore Target Computer
Real-time application and multicore target computer configuration for concurrent execution