Nuvoton unveils advanced Endpoint AI Platform, introducing cutting-edge MCU and MPU silicon for streamlined machine learning development.
Nuvoton recently unveiled its Endpoint AI Platform, a robust system designed to expedite the development of sophisticated microcontroller (MCU) AI products. This platform harnesses Nuvoton’s latest MCU and MPU silicon technology, featuring the NuMicro M55M1 with Ethos U55 NPU, NuMicro MA35D1, and NuMicro M467 series.
These MCUs add significant value to the AI-centric computing toolkit and underscore Nuvoton’s collaboration with industry leaders to create a comprehensive Endpoint AI Ecosystem.
The development on these platforms is facilitated by Nuvoton’s NuEdgeWise, a versatile machine learning (ML) tool suitable for various ML applications. This amalgamation of hardware and development tools solidifies Nuvoton’s position as a leading microcontroller platform provider.
These single-chip platforms find application in smart home appliances, security, smart city services, industry, agriculture, entertainment, environmental protection, education, voice-controlled tasks, sports, health, and fitness.
NuMicro M55M1 Features:
The NuMicro M55M1 series targets machine learning applications with Ethos-U55 NPU and on-device AI capabilities, ideal for embedded applications. Its low-power monitoring ability through image sensors, microphones, and sensors without waking the central processor is noteworthy.
Additionally, the M55M1 MCU includes an ML model protection mechanism, enhancing security against potential cyber threats. Supporting Arm Helium technology, these processors offer improved performance for ML and DSP applications in compact embedded systems.
NuMicro MA35D1 Capabilities:
The MA35D1 series, a heterogeneous multi-core microprocessor for Edge IIoT Gateway applications, features a dual-core 64-bit Arm Cortex-A35 core at 800 MHz and a 180 MHz Arm Cortex-M4, enabling Tiny AI/ML edge computing.
Versatile M467 MCU:
The M467 series, a 32-bit microcontroller built on the Arm Cortex-M4F core with DSP instruction sets and single precision floating-point unit (FPU), caters to diverse applications—from smart home appliances to IoT gateways, industrial control, telecommunications, and data centers.
In IoT, the M467 integrates various connectivity, I/O, and security peripherals, allowing tailored hardware extensions for specific applications. The M467’s HyperRAM offers memory flexibility for different ML models while ensuring power efficiency and bandwidth adaptability.
Comprehensive Development Support:
Nuvoton provides fully-featured development boards and robust tools. For example, AI application development using NuMaker MA35D1 supports machine learning projects and user analysis via the Human Machine Interface (HMI). The NuMaker-IoT-M467 board is tailored for M467 MCU IoT applications.
NuEdgeWise ML IDE for Simplified TinyML Development:
NuEdgeWise IDE streamlines TinyML development, supporting critical ML application stages through TensorFlow Lite deployment on Nuvoton chips via the Jupyter Notebook platform, simplifying TinyML implementations.
Hi…Just have a look at this also!