Support for
Model-Based Design
(MBD)

Simulink Performance Estimator icon

Get to Know the Runtime of Your Models

Function developers who use model-based design (MBD) often don’t know how their design decisions will affect the application’s runtime on the target embedded platform. Decisions about which data types to use or how to implement functionality can have a significant impact depending on the target platform.

In many workflows, models are converted to C code using tools such as TargetLink, Embedded Coder, ASCET, or SCADE. emmtrix Performance Estimator then automatically analyzes the expected runtime behavior on the selected embedded target. Using various mapping strategies, the results can be mapped back to the original models and subsystems.

This feature allows function developers to obtain information about the runtime impact on the selected embedded platform when selecting realizations or setting options. The results are presented in an HTML report with features such as sorting, filtering, and colored indicators about estimation confidence.

emmtrix Performance Estimator Workflow
with TargetLink/Embedded Coder

Figure 1 shows our emmtrix Performance Estimator workflow for Simulink models and Stateflow charts.

emmtrix Performance Estimator Workflow

Figure 1: emmtrix Performance Estimator with TargetLink/Embedded Coder Support

Key Features

  • Automatic Reports and Visualization: Automatically generate reports and visualizations with detailed performance information.
  • Confidence Levels: Classify estimation results using confidence levels.
  • Workflow Integration: Integrate performance estimation into your development workflow.
  • Block Execution Times: Obtain minimum and maximum execution times of blocks.
  • Interactive HTML Reports: Filter, search, and sort the results.

Key Benefits

  • Early Performance Insights: Assess performance during model-based development, early in the development process.

  • Continuous Monitoring: Track performance changes throughout development with CI flow integration.

  • Better Timing Understanding: Gain detailed insights into the timing behavior of your application.

  • Focused Optimization: Identify high-runners and critical parts of your models.

  • Preparation for Hardware Tests: Ensure your code is ready for hardware tests.

  • No Instrumentation Overhead: Obtain runtimes for subsystems and modules without instrumenting the source code.

Try the Interactive Performance Report

Explore the interactive HTML example report below. Use the filters, search, and sorting options to inspect the results and expand the hierarchy to explore individual model elements and subsystems.

Figure 2: HTML Report of Results

Use Cases

Identifying Bottlenecks in MBD Function Development

Identify performance bottlenecks in generated code and relate them to Simulink blocks or subsystems.

Technical background and application

In Model-Based Design (MBD), software functions are typically developed in tools like Simulink before being automatically translated into C code. However, performance bottlenecks often arise from inefficiencies in auto-generated code. A static performance analysis tool can identify such inefficiencies at an early stage by analyzing the generated code structure and linking performance issues to Simulink blocks or subsystems. This allows function developers to pinpoint modeling inefficiencies, optimize computationally expensive constructs, and align the generated code with performance expectations. By integrating this analysis into the development workflow, potential execution overheads can be detected before deployment, reducing debugging efforts and improving the efficiency of MBD-based software development.

Enhancing MBD Function Developers’ Performance Awareness

A static performance analysis tool can provide immediate feedback by analyzing the generated code and highlighting performance-relevant design choices.

Technical background and application

Model-Based Design (MBD) abstracts low-level implementation details, making it easier for function developers to focus on algorithms rather than code efficiency. However, this abstraction often leads to a lack of awareness regarding the performance impact of specific Simulink constructs or model settings.

A static performance analysis tool can provide immediate feedback by analyzing the generated code and highlighting performance-relevant design choices. By linking performance insights to Simulink blocks and subsystems, developers gain a deeper understanding of how their modeling decisions affect execution time and computational efficiency. This continuous feedback loop enables them to develop an intuition for performance trade-offs, ultimately leading to more optimized and efficient model designs before deployment.

Assessing the Impact of MBD Design Decisions on Performance

A static performance analysis tool enables developers to track the impact of key modeling decisions, such as using Stateflow vs. Block-Based Modeling or applying Signal Vectorization.

Technical background and application

In Model-Based Design (MBD), different modeling approaches can lead to significant performance variations in the generated C code. However, function developers often lack visibility into how specific design choices affect execution efficiency.

A static performance analysis tool enables developers to track the impact of key modeling decisions, such as using Stateflow vs. Block-Based Modeling or applying Signal Vectorization. By continuously analyzing the generated code, developers can compare performance characteristics across design iterations and assess whether certain modeling approaches lead to more efficient execution. This systematic tracking helps prevent unintended performance regressions and supports informed decision-making to optimize software before integration and testing.

Does one of these use cases sound familiar? Let’s talk about your specific scenario.

Some Supported Platforms

A wide range of supported target platforms ranging from general-purpose processors (e.g. ARM Cortex-A series or X86) to special-purpose microcontrollers (e.g. Infineon AURIX™ family) are already supported. In general, the performance estimation can be easily adapted and customized to provide basic support for new processor architectures. More complex and accurate hardware models with respect to the processor pipeline can be supported on demand.

Infineon logo

AURIX™
(TC2xx, TC3xx, TC4x)

arm logo

Cortex-A
(e.g. A55, A76)

Power PC logo

PowerPC

Raspberry Pi logo

Raspberry Pi

Renesas logo

R-Car

Samsung logo

Samsung

Associated Partners

emmtrix Technologies is an Infineon Associated Partner with over 10 years of experience working with the Infineon AURIX™ microcontroller family and has been actively collaborating with Infineon for the past five years.

emmtrix Technologies is a member of the MathWorks Connections Program and offers emmtrix Performance Estimator for early, reproducible runtime insights into C code generated from Simulink® models for embedded targets.

Technical Resources

Technical White Paper

Why continuous monitoring of software timing is key in complex software projects.

I

Webinar Recording

Runtime of Simulink Models – How to get a Grip

From Model to Metrics

Building embedded systems from models is a great start – but what about timing, memory and platform constraints?
emmtrix Performance Estimator (ePE) helps you understand how your generated code behaves before execution – statically, accurately, and CI-ready.

Frequently Asked Questions

Does ePE analyze the model or the generated C code?

ePE analyzes the generated C code to estimate its expected runtime behavior on the selected embedded target. Using mapping strategies, the results can be mapped back to the original models and subsystems.

Can ePE be used with TargetLink and Embedded Coder?

Yes, both code generators are supported. The TargetLink integration has been successfully used in several automotive use cases, while Embedded Coder support is currently more limited. Contact emmtrix to discuss the available model mapping options for your specific toolchain.

Which modeling decisions can affect runtime?

Data types, the implementation of functionality, model constructs, and model settings can affect runtime on the target embedded platform. Analysis can help compare choices such as Stateflow versus block-based modeling or signal vectorization.

How are the results presented?

The results are presented in an interactive HTML report with filtering, searching, sorting, and color-coded indicators of estimation confidence. Results can be linked to models and subsystems. Additionally, machine-readable formats such as JSON and XML are available for integration into automated workflows and external tools.

Can performance changes be monitored during development?

Yes. The workflow supports continuous monitoring of performance changes during development with CI flow integration. Repeated analysis helps developers compare performance characteristics across design iterations.

Does the workflow require source-code instrumentation?

The described workflow provides runtimes for subsystems and modules without instrumentation of the source code.

Let’s Get in Touch

Would you like to understand the timing impact of your modeling decisions? Contact us to discuss your modeling environment, code generator, target platform, and timing requirements, or request a demo.

Cookie Consent with Real Cookie Banner