MCU+AI Accelerators Becoming Increasingly Popular

According to Yole Intelligence, part of Yole Group, in the Microcontroller (MCU) Market Monitor, despite the significant surge in global inflation during 2022 and numerous warnings from financial institutions about a potential recession, the MCU market thrived in that year. It experienced an impressive annual revenue growth of 25%.Although many suppliers had concerns about overstock earlier in the year, the demand surpassed expectations, leading to substantial growth in the second half of the year, before the impact of inflation became pronounced.

In contrast to the rest of the semiconductor industry, which faced challenges such as a sharp decline in consumer spending and oversupply issues resulting in lower average selling prices and revenues, the MCU market resisted price drops and maintained levels significantly higher than those seen before the pandemic. It is anticipated to continue this upward trajectory, aiming for a flat growth rate in 2023. The prospects for 2024 also indicate low-single digit growth, while the long-term outlook for the MCU market appears particularly promising.

The Microcontroller (MCU) Market is projected to have remarkable growth, with a Compound Annual Growth Rate (CAGR) of 6.97% between 2022 and 2027. This expansion is expected to result in a substantial increase in market size, amounting to an impressive USD 9.03 billion growth. A comprehensive analysis of the market reveals that the increasing use of microcontrollers in automobiles, increasing demand for IoT and smart devices, and the growth of the consumer electronics industry are the primary driving forces behind this robust growth.

During this process, there has been a continuous growth of interconnected devices and a trend towards developing more intelligent devices at the edge of the Internet of Things (IoT). As AI penetrates into the edge and terminal devices, the development of edge computing has become one of the major trends driving the global MCU market growth.

These trends not only drive the market demand for MCUs but also foster new technological innovations in MCUs.

It is well known that previous AI computations and machine learning were primarily done in the cloud, but now they are gradually shifting towards the edge. Edge computing refers to computations performed at the network edge, closer to the source of data generation, rather than in remote data centers. The benefit of this approach is to reduce bandwidth usage for cloud uploads, improve response time of local devices, and enhance the security of local data.

With the continuous deployment of edge computing in the IoT field, the use of MCUs in edge computing will increase. Meanwhile, in the era of IoT where everything is connected, data is growing exponentially, putting immense computational pressure on CPUs. Various solutions have emerged in the market to address "how to alleviate CPU computational pressure." Some companies have started incorporating accelerators in MCUs to perform ML computations using dedicated processing power, aiming to release the general computational power of CPUs.

In short, AI and ML computations are migrating from the cloud to the edge. Prioritizing computation at the edge can improve the responsiveness of local devices, reduce bandwidth usage for cloud uploads, and enhance the security of local data, among other benefits. This edge AI trend will not be limited to smartphones or computers with SoC-level computing power, but will extend further to IoT devices predominantly controlled by MCUs.

In response, MCU manufacturers are innovating and attempting to integrate AI/ML technologies into MCUs to keep pace with smarter edge devices.

【Source: Semiconductor Industry Observation】