Ambiq Micro, Inc. is an American semiconductor company specializing in ultra-low-power integrated circuits. Based in Austin, Texas, it develops microcontrollers and systems on a chip for devices with limited energy resources. Its components are used notably in smartwatches, wearable accessories, sensors and certain health monitoring devices.
From university research to the connected device market
Founded in 2010, Ambiq grew out of research conducted at the University of Michigan on energy-efficient electronic circuits. Scott Hanson, David Blaauw and Dennis Sylvester are among its founders. The company was built around a technical approach: substantially reducing the operating voltage of circuits to limit their power consumption while maintaining performance suited to embedded applications.
This approach is marketed under the name SPOT, which stands for “Subthreshold Power Optimized Technology”. It notably involves operating transistors at voltages close to or below their threshold voltage. The challenge lies in managing the speed, reliability and variability constraints associated with these operating conditions.
Components for processing data as close to the sensor as possible
The Apollo family is a central part of Ambiq’s offering. It includes microcontrollers and systems on a chip based on Arm cores, with configurations tailored to different computing, memory and connectivity requirements. Some models incorporate Bluetooth Low Energy communication, which is useful for connecting a wearable device to a smartphone or a gateway.
The benefits of these components go beyond keeping a device in standby mode for longer. They also involve reducing the energy required for repeated tasks: taking measurements, analyzing a signal, displaying an interface or recognizing an event. In wearable devices, this energy efficiency can help extend battery life or allow the use of a smaller battery, depending on the manufacturer’s choices.
Ambiq provides software tools alongside its chips for embedded development, particularly for running artificial intelligence models locally. The company follows a fabless model: it designs its circuits and relies on industrial partners to manufacture them. It therefore serves equipment manufacturers rather than end consumers.
What’s next?
The development of embedded AI is broadening the range of possible applications for Ambiq. Processing audio signals or motion data directly can limit transmissions to the cloud, reduce latency and, depending on the architecture chosen, better protect certain personal data.
However, the company’s progress will depend on practical trade-offs: actual system power consumption, computing power, component costs and ease of integration. When competing with other suppliers of low-power microcontrollers, the quality of software tools and support for designers will matter as much as the chips’ electrical characteristics.