Beijing, China, May 6, 2020 – MathWorks today launched version R2020a, extending AI capabilities for deep learning. Engineers can now train neural networks in the updated Deep Network Designer app, manage multiple deep learning experiments in the new Experiment Manager app, and choose a method for generating deep learning code from more network options. Not only does R2020a introduce hundreds of new and updated features for all MATLAB and Simulink users, it also introduces new features specifically for automotive and radio engineers. For more details, watch the R2020a video.
AI and Deep Learning
“MathWorks provides a comprehensive platform for building AI-driven systems,” said David Rich, Director of Marketing at MATLAB. “We have built more than three decades of product, consulting and support experience into AI workflows that empower engineers and scientists Ability to clean data, build models, and deploy those data and models in production IT or embedded systems.”
The new release introduces an enhanced Deep Learning Toolbox that helps users:
Use the new Experiment Manager app to manage multiple deep learning experiments, track training parameters, analyze and compare results and code.
Use the Deep Network Designer app to interactively train networks for image classification, generate MATLAB code for training, and access pretrained models.
Additionally, GPU Coder now offers a broader set of networks to implement AI systems on cloud and edge devices, including Darknet-19, Darknet-53, Inception-ResNet-v2, NASNet-Large, and NASNet-Mobile.
New Products, Important Updates, and Enhanced Product Features
R2020a includes three new products:
Motor Control Blockset includes a library of motor control algorithms optimized for generating compact code and provides out-of-the-box support for multiple motor control hardware kits.
Simulink Compiler enables engineers to generate stand-alone applications, web applications, and software components from Simulink models and run simulations without installing Simulink.
MATLAB Web App Server provides users with controlled browser access to MATLAB web applications deployed across the organization.
The new version also includes the following important updates:
Automotive engineers can more easily create driving scenarios using road data imported from high-definition geographic maps and optimize shift schedules for performance, fuel economy, and emissions analysis.
For radio engineers, MATLAB support for 5G and WiFi-6 technologies has been extended, adding support for waveform generation and cell detection.
Users can generate C++ classes from MATLAB classes, and the new version introduces message-based communication for service functions and reusable subsystems.
R2020a includes hundreds of other enhancements to the MATLAB and Simulink product families, including:
MATLAB now has improved support for non-ASCII character sets with default UTF-8 encoding, new graph customization features and new chart types for visualizing data, and for interactive retiming or synchronization of timetables and stacks or non- Live editor tasks for stack table variables.
Simulink users can now write C code directly into their models to integrate external algorithms, use the new C function block to easily add pre-configured C functions for preprocessing or postprocessing, and specify different code for simulation and code generation.
The R2020a is now available globally. For information on all new products, enhancements, and bug fixes for the MATLAB and Simulink product families, visit the R2020a Release Highlights page.
Graphic: MathWorks’ R2020a helps engineers interactively train an image network
MathWorks is a leading developer of mathematical computing software. It launched MATLAB, known as the “language of scientists and engineers,” and is a programming environment for algorithm development, data analysis, visualization, and numerical computation. Simulink is a block diagram environment for multi-domain and embedded engineering system simulation and model-based design. Engineers and scientists around the world rely on these product lines from The MathWorks Corporation to accelerate the pace of invention, innovation and development in the automotive, aerospace, electronics, financial services, biomedical and other industries. MATLAB and Simulink products are also essential teaching and research tools at universities and academic institutions around the world. MathWorks was founded in 1984 and is headquartered in Natick, Massachusetts, with more than 4,500 employees in 16 countries.