Nuvoton has introduced the Endpoint AI artificial intelligence platform to accelerate the development of full-featured AI microcontroller (MCU) products. These solutions are enabled by Nuvoton’s powerful new MCU and MPU silicon, including the NuMicro® M55M1 microcontroller, equipped with Ethos U55 NPU accelerator, NuMicro® MA35D1, and NuMicro® M467 series. These MCUs are a valuable addition to the modern AI-centric computing toolkit and demonstrate how Nuvoton continues to work closely with Arm and other companies to develop a user-friendly and complete Endpoint AI Ecosystem.
Most important features of the Endpoint AI Platform:
- AI and ML single-chip products,
- the NuEzAI-M55M1 development board, powered by the NuMicro® M55M1 microcontroller, which is based on the Arm® Cortex®-M55 core,
- compact design and low power consumption for machine learning,
- designed for today’s and tomorrow’s ML applications,
- backed by a deep development environment and toolkit,
- equipped with Nuvoton’s NuEdgeWise software – a well-rounded, simple-to-adopt tool for machine learning (ML) development, which is nonetheless suitable for cutting-edge tasks.
These new single-chip-based platforms are ideal for applications including TinyML (executing ML models directly in embedded systems), Industrial IoT (IIoT), smart home, smart city, environmental protection, education, services, highly accurate voice-control tasks, sports, health, and more.
New ML-Focused Hardware with NPU: NuMicro M55M1
The NuMicro M55M1 series microcontroller possesses abundant computational resources, including DSP extensions, vector extensions, floating-point unit (FPU), and Neural Network (NN) accelerator, as well as up to 1.5 MB of SRAM and 2 MB of flash memory, making it an ideal choice for embedded applications. It is particularly suitable for processing-intensive workloads.
The NuMicro M55M1 MCU supports secure boot, TrustZone, True Random Number Generator (TRNG), key storage, cryptographic accelerators, tamper detection, and PSA certification Level3 compliance, significantly enhancing the security and integrity of systems built with the M55M1.
In terms of analog and control functions features 2 sets of 12-bit 5MSPS SAR ADC, 2 sets of 12-bit 1MSPS buffered DAC, 4 analog comparators, 1 built-in temperature sensor, 1 internal voltage reference module, up to 24-channel 200 MHz PWM output, 4 QEI, and 4 input capture units.
The M55M1 series also offers a rich set of communication and peripheral interfaces, including Universal Serial Control Interface (USCI) configurable as UART/SPI/I2C, 10 sets of UART, 4 sets of SPI/I2S, 4 sets of I2C, 2 sets of CAN FD, 2 sets of QSPI, USB 2.0 HS OTG/host/device, USB 2.0 FS OTG/host/device, 10/100 Ethernet MAC with RMII, 8-bit parallel camera interface (CCAP), 2 PDM with internal filters, and 2 SDIO.
With its rich features, the NuMicro M55M1 series microcontroller is targeted at machine learning (ML) systems. It is supported by the built-in Ethos-U55 NPU (Neural Processing Unit) and implemented AI functions suitable for embedded applications. It allows the system – based on the image sensor, microphone and other sensors – watch for events while in low-power mode, without waking up the central processor. The M55M1 microcontroller includes an ML model protection mechanism that enhances security by safeguarding ML intellectual property against potential malicious hacking attempts. This is one of the first processors to support Arm Helium technology, which provides a significant performance boost for machine learning (ML) and digital signal processing (DSP) applications in small, low-power embedded systems.
Edge IIoT Gateway Solution: NuMicro MA35D1
The MA35D1 series is a heterogeneous multi-core microprocessor for high-end Edge IIoT Gateway, based on a dual-core 64-bit Arm
Cortex-A35 core at 800 MHz and a 180 MHz Arm Cortex-M4. These high-performance cores facilitate Tiny AI/ML edge computing.
The Versatile M467, with IoT Applications and Great Connectivity
The M467 series is a 32-bit microcontroller based on the Arm Cortex-M4F core with a built-in DSP instruction set and single precision floating point unit (FPU). It is ideal for a wide range of applications: smart home appliances, IoT gateways, industrial control, 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. With the M467’s broad built-in I/O support, users can choose only the precise hardware extensions they need for their particular applications. The M467 also supports HyperRAM. In AI/ML applications, the 64MB of HyperRAM provides the flexibility to handle different ML models with varying memory size or density requirements.
Strong Development Support
Fully functional development boards are available for all the above hardware applications. They are supported by Nuvoton’s advanced development tools and development environment. Below are two of them:
NuMaker-M55M1 development board to help design various solutions using the NuMicro M55M1 series microcontroller.
The NuEzAI-M55M1 development board to support the design of various solutions for Machine Learning and the Endpoint AI platform.
NuEzAI-M55M1 is a development board equipped with NuMicro® M55M1 series microcontroller, offering powerful digital signal processing and on-device machine learning inference capabilities. Powered by Arm® Cortex®-M55 core, M55M1 microcontroller runs at a frequency of up to 220 MHz. Additionally, integrates Arm® Ethos™-U55 Neural Processing Unit (NPU) as a coprocessor, significantly enhancing neural network computation performance. The M55M1 series MCU features up to 1.5 MB of SRAM and 2 MB of flash memory, providing ample storage for applications.
It offers a wide range of hardware expansion options, including CCAP, DMIC, G-Sensor, HSUSB Type-C, HyperRAM, Nu-Link2-Me, MicroSD, and an Arduino-compatible expansion interface, allowing users to customize functionality as needed.
The NuEzAI-M55M1 board requires no driver installation and seamlessly integrates with the Google Teachable Machine platform, which is an intuitive, no-code, visual platform that enables users to easily train and export AI models.
NuEdgeWise ML IDE
Nuvoton’s NuEdgeWise IDE (Integrated Development Environment) is a machine-learning tool designed for TinyML development. The IDE supports the four key stages of ML application development: labeling, training, validation, and testing. NuEdgeWise leverages the popular Jupyter Notebook platform, allowing developers to train and deploy models on Nuvoton chips using TensorFlow Lite. This makes TinyML applications more accessible and easier to implement.
