© ROOT-NATION.com - Use of content is permitted with a backlink.
Anker Innovations has officially presented Thus, a proprietary AI platform based on a chip designed to improve sound quality. The production is set up in Germany, and the debut of new soundcore devices with this chip is scheduled for the event on May 21, 2026.
Thus is Anker’s proprietary AI platform that enables localized neural network processing in consumer electronics. Initially, the company plans to implement this processor in its audio line, and later extend it to mobile accessories and IoT devices.
The architecture of Thus is fundamentally different from the traditional serial approach, in which memory and computing units are physically separated. Instead, Anker uses compute-in-memory (CIM) technology, which eliminates the energy-intensive movement of data between components. This solution is especially useful for compact devices such as in-ear headphones, where battery life is extremely limited.
Modern top-of-the-line headphones usually rely on small neural networks with a limited number of parameters due to strict power consumption requirements. Anker, however, assures that the Thus chip is capable of supporting millions of parameters and provides 150 times more processing power for noise reduction compared to previous soundcore models. In particular, the Clear Calls feature utilizes a large-scale neural network on the device, along with eight microphones and bone conduction sensors to isolate voices in noisy environments.
The name Thus comes from the phrase “Thus have I heard,” a reference to the direct transfer of knowledge. The company sees this chip as a long-term platform, not a one-off product. More detailed specifications and device-specific announcements are expected at the Anker Day event in New York on May 21.
Read also: