From the series: Coder Summit
Andy Bartlett, MathWorks
When your model and generated code results do not match, denormal numbers may be the cause.
Flush-to-zero math, in particular, is a common cause for mismatches since gradual underflows used in simulation are not often used in embedded systems.Most embedded design uses Flush to zero for faster math.
Learn how new features in Fixed-Point Designer™ let you use gradual underflow for plant of environment models, and Flush to zero for your embedded algorithms.
This approach provides the best numerical accuracy for system simulation, while providing more optimized code that matches your embedded system.
Concurrent Execution of For-Loops Hear how Embedded Coder supports concurrent, parallel, and multi-thread execution of for-loops.
Dynamic Code Verification and Analysis: SIL, PIL, and External Mode See how you can use Embedded Coder to verify, tune, and log generated code using SIL, PIL, and External mode.
C-Code Generation and Integration with C# Learn how you can use MATLAB Coder and Simulink Coder to integrate with a C# application.
Introducing AUTOSAR Blockset for Classic and Adaptive Applications Learn how to design, simulate, and generate code for AUTOSAR Adaptive and Classic applications.
Simulate with Flush to Zero to Avoid Pains in Model-Based Design Workflows Learn how to simulate and generate code for denormal floating-point numbers using Flush to zero.
Embedded Code Generation Advances Hear about embedded code generation advances in R2018a with the major release of MATLAB Coder and Embedded Coder.
How to Generate Production Code in 5 Minutes Watch how to quickly customize and generate code using Embedded Coder Quick Start, Code Perspective, and Embedded Coder Dictionary.
Embedded Coder Data, Function, and File Customization Learn about recent improvements to Embedded Coder that let you customize generated data, functions, and files according to your production software environment and standards.
Optimizing Code Made Easy Using Embedded Coder See how Embedded Coder lets you quickly use specific optimization levels in R2018a and how it has a multifaceted approach to producing more efficient code.
Code Generation Improvements for Vision and ADAS Watch how MathWorks coder products, such as MATLAB Coder and Embedded Coder share a common coder engine that is increasingly optimized for code patterns in computer vision and automated driving/ADAS applications.
User-Suggested Data Reuse for More Efficient Code Hear how Embedded Coder supports automatic data buffer reuse, user-specified reuse, and starting in R2018a, an intuitive user-suggested data reuse option that minimizes unnecessary data copies in generated code.
SIMD Code Generation Hear how Embedded Coder generates native SIMD instructions including Intel SSE and AVX for Windows and Linux. Previous SIMD support used code wrappers but native SIMD generation in R2018a improves quality and efficiency.
ADAS Design, Simulation, and Code Generation with Simulink See how Automated Driving Toolbox and Embedded Coder let you design, simulate, and generate C++ code for automated driving and ADAS systems. Examples include adaptive cruise control with sensor fusion and model predictive control.
C++ Class Modeling and Code Generation with Simulink and Embedded Coder Hear how Simulink and Embedded Coder let you model classes and generate C++ code. Object-oriented design and realization is critical for services-oriented architecture and middleware including Adaptive AUTOSAR, ROS 2.0, DDS, and more.
Optimizing Lookup Tables in Simulink and Embedded Coder Fixed-Point Designer lets you optimize lookup tables in Simulink and generate highly efficient code with Embedded Coder. The Lookup Table optimizer compresses data into data types that minimize memory use with low impact to numerical accuracy.
Choose a web site to get translated content where available and see local events and offers. Based on your location, we recommend that you select: .Select web site
You can also select a web site from the following list:
Select the China site (in Chinese or English) for best site performance. Other MathWorks country sites are not optimized for visits from your location.