Analog Devices' MAX78000 microcontroller has been recognized for its innovative design and contribution to the field of edge technology by being named a winner of a 2024 Best of Sensors Award. The awards, co-hosted by Fierce Electronics and Sensors Converge, celebrate the most groundbreaking advancements in sensor technologies. The MAX78000's triumph in the Edge Technology category highlights its potential to shape the future of sensor data processing at the network's periphery.
The MAX78000 microcontroller from Analog Devices combines processing capabilities with ultra-low power consumption. This system-on-a-chip (SoC) enables artificial intelligence (AI) and machine learning applications to run efficiently on battery-powered devices at the edge of Internet of Things (IoT).
At the core of the MAX78000 Is a neural network accelerator, specifically optimized for convolutional neural networks (CNNs). This hardware-based CNN engine has 442 KB of weight storage memory, supporting 1-, 2-, 4-, and 8-bit weights. This flexibility allows for networks of up to 3.5 million weights, providing capacity for complex AI models. The SRAM-based CNN weight memory enables on-the-fly AI Network updates, enhancing adaptability for real-world applications.
Complementing the CNN accelerator is an Arm® Cortex®-M4 Processor with a floating point unit (FPU), capable of operating at speeds of up to 100 MHz. This processor handles system control and general-purpose computing tasks efficiently. The MAX78000 also incorporates a 32-bit RISC-V coprocessor running at up to 60 MHz, further adding to its computational capabilities.
The device's memory architecture features 512 KB of flash memory and up to 128 KB of SRAM for the microprocessor core. This allocation of on-chip memory ensures the smooth operation of complex algorithms and data processing tasks. The inclusion of a 16 KB instruction cache optimizes performance, while the optional Error Correction Code (ECC-SEC-DED) for SRAM enhances data integrity.
The chip's power management system is designed to maximize battery life in portable applications. It incorporates a Single-Inductor Multiple-Output (SIMO) switch-mode power supply, operating within a 2.0 V to 3.6 V range. Dynamic voltage scaling further minimizes active core power consumption, resulting in a low power draw of just 22.2μA/MHz during loop execution at 3 V from the cache.
The MAX78000's versatility is evident in its wide variety of potential applications. It excels in tasks such as object detection and classification, audio processing (including multi-keyword recognition and sound classification), facial recognition, and time series data analysis for applications like health monitoring and predictive maintenance.
To facilitate integration into various IoT ecosystems, the MAX78000 supports multiple high-speed and low-power communication interfaces. These include I2S for digital audio and a parallel camera interface (PCIF), making it well-suited for image and video processing applications.
Security features were also a consideration in the MAX78000's design. The chip includes an AES 128/192/256 hardware acceleration engine and a True Random Number Generator (TRNG) seed generator, addressing the need for data security in IoT devices.
The MAX78000's ability to perform AI inference while consuming only microjoules of energy sets it apart in the field of edge computing. This efficiency allows for the deployment of sophisticated AI applications and scenarios where power consumption is a critical constraint, such as in wearable devices, remote sensors, and battery-operated IoT nodes.
Analog Devices, Inc. has been a leader and innovator in the semiconductor industry for over five decades. Founded in 1965 by MIT graduates Ray Stata and Matthew Lorber, the company has grown to serve approximately 100,000 customers across various sectors, including communications, computing, instrumentation, and consumer electronics. Analog Devices' expertise spans analog signal processing, digital signal processing, data converters, amplifiers, and more. This recognition for the MAX78000 Microcontroller furthers Analog Devices' position as a leader in edge computing solutions and its commitment to addressing the evolving needs of the electronics industry.
To discover more about the innovations honored at this year's event, including winners in other categories, visit this article.