TL;DR:
- Nuvoton introduces Endpoint AI Platform with advanced MCUs and MPUs.
- NuEdgeWise simplifies machine learning (ML) development for various applications.
- NuMicro M55M1 with Ethos-U55 NPU enables low-power event monitoring.
- MA35D1 series offers high-end Edge IIoT Gateway solutions.
- M467 series provides versatile 32-bit microcontroller options.
- HyperRAM support enhances memory flexibility.
- Fully-featured development boards are available for all hardware applications.
- NuEdgeWise IDE simplifies TinyML development using TensorFlow Lite.
Main AI News:
Nuvoton has introduced its latest Endpoint AI Platform, revolutionizing the landscape of microcontroller (MCU) AI products. This innovative platform leverages Nuvoton’s cutting-edge MCU and MPU silicon technology, featuring the NuMicro® M55M1 equipped with Ethos U55 NPU, NuMicro® MA35D1, and NuMicro® M467 series. These MCUs represent a significant step forward in AI-centric computing, highlighting Nuvoton’s ongoing collaboration with industry leaders like Arm to deliver a comprehensive and user-friendly Endpoint AI Ecosystem.
What sets these platforms apart is Nuvoton’s NuEdgeWise, a versatile and user-friendly tool for machine learning (ML) development. While accommodating advanced tasks, NuEdgeWise simplifies the development process, making it accessible even for cutting-edge applications. Coupled with robust development tools, these single-chip-based platforms solidify Nuvoton’s position as a leading microcontroller platform provider. These platforms find their ideal applications in smart home appliances, security systems, smart city services, industrial solutions, agriculture, entertainment, environmental protection, education, highly precise voice-control tasks, sports, health, and fitness.
Introducing the NuMicro M55M1 with NPU
The NuMicro M55M1 series microcontroller is tailored for machine learning applications, thanks to its Ethos-U55 NPU (Neural Processing Unit) and on-device AI capabilities, which are perfect for embedded applications. This MCU empowers the system to monitor events – whether triggered by image sensors, microphones, or other sensors – while conserving power in low-power mode, without the need to wake up the central processor. Notably, the M55M1 MCU includes an ML model protection mechanism, enhancing security by safeguarding ML intellectual property against potential malicious attacks. These processors are among the pioneers to support Arm Helium technology, offering a substantial performance boost for machine learning (ML) and digital signal processing (DSP) applications in compact, energy-efficient embedded systems.
The MA35D1 series boasts a heterogeneous multi-core microprocessor designed for high-end Edge IIoT Gateway applications. It features a dual-core 64-bit Arm Cortex-A35 core operating at 800 MHz and a 180 MHz Arm Cortex-M4, enabling robust Tiny AI/ML edge computing capabilities.
Unveiling the M467 series
The M467 series is a 32-bit microcontroller built around the Arm Cortex-M4F core, featuring a built-in DSP instruction set and single precision floating-point unit (FPU). Its versatility makes it ideal for a wide range of applications, including smart home appliances, IoT gateways, industrial control systems, telecommunications, and data centers. In IoT tasks, the M467 can be enhanced with a rich set of connectivity, I/O, and security peripherals, from Ethernet 10/100 MAC to hardware encryption, decryption, and key storage. The M467’s extensive built-in I/O support allows users to customize their hardware extensions for specific applications. Additionally, the M467 offers support for HyperRAM, which, in AI/ML applications, provides 64MB of flexibility to handle various ML models with varying memory size or density requirements. HyperRAM also offers power-saving features, compatibility with available bandwidth, ease of use, and the option for flexible memory expansion.
Complete Development Ecosystem
Nuvoton’s commitment to developers is further evident with fully-featured development boards available for all the aforementioned hardware applications. These development boards are backed by Nuvoton’s extensive development tools, environment, and dedicated support. For instance, AI application development with NuMaker MA35D1 enables not only efficient machine learning projects, such as image classification but also presents intuitive analysis to users through the Human Machine Interface (HMI). Meanwhile, the NuMaker-IoT-M467 development board is tailored for IoT applications involving the M467 MCU.
NuEdgeWise IDE
Simplifying TinyML Development Nuvoton’s NuEdgeWise Integrated Development Environment (IDE) is a powerful machine-learning tool designed for TinyML development. This IDE supports the four key stages of ML application development: labeling, training, validation, and testing. Leveraging the popular Jupyter Notebook platform, NuEdgeWise empowers developers to train and deploy models on Nuvoton chips using TensorFlow Lite. This breakthrough ensures that TinyML applications are more accessible and easier to implement than ever before.
Conclusion:
Nuvoton’s unveiling of the Endpoint AI Platform marks a significant advancement in MCU AI products. With a comprehensive lineup of MCUs, including the power-efficient M55M1, high-performance MA35D1, and versatile M467 series, coupled with the user-friendly NuEdgeWise development tool, Nuvoton solidifies its position as a leader in microcontroller platforms. This innovation has the potential to reshape the market by making advanced AI capabilities more accessible across various industries, from smart homes to IoT and industrial applications, promising a brighter future for technology.
