In line with the mega trend of intelligence and edge AI applications, Taiwan's MCU makers are focusing on the development of MCUs integrated with AI to explore the AI MCU business opportunities and create more new edge AI markets and applications, according to industry sources.
The sources said Taiwanese MCU suppliers have long been engaged on consumer electronics applications, but they are finding opportunities in the edge AI market as consumer electronics devices are gradually incorporating AI functions to go intelligent. However, high complexity of AI chip design and the need for big data analysis and deep learning algorithms are making it more difficult to enter the AI application market, the sources continued.
Among the suppliers, Holtek Semiconductor is set to release diverse module solutions combining MCUs and sensors in 2023. This strategy is not only to enhance its product differentiations but also to boost the added value of MCUs in response to market trends and needs, according to company sources.
Holtek noted it is also developing products associated with edge AI computing, such as different specs of blood oximeters and health measurement devices that have different pressure, speed and flow requirements.
Another MCU maker Nuvoton Technology has years of deployments in the edge AI field, focusing on industrial applications such as industrial noise detection, which can be improved to a certain extent through AI MCUs.
Nuvoton said that AI server applications are gaining traction and will increase sharply in the next few years, which will positively contribute to its overall business operations. The company and Microsoft are jointly developing a remote server BMC (baseboard management controller) solution, which is expected to start revenue contribution in 2024.
Nuvoton pointed out that customers are showing ever-higher requirements for information security, and therefore the company and Microsoft will continue to assess the feasibility of developing other AI-related application markets.
Industry sources also said that edge AI computing can reduce data latency and save the cost of cloud computing. By combining MCU with edge AI applications, devices can be processed directly at the front-end without the need to be sent to the cloud, promising new business opportunities for Taiwan MCU makers if they keep advancing to the segment of AI MCUs, the sources stressed.