Anker has unveiled its proprietary "Thus" chip, designed to embed powerful, localized artificial intelligence capabilities directly into small, power-constrained hardware like headphones and other wearables. This move aims to offload complex AI processing from the cloud, keeping computations entirely on the device itself. The chip’s debut is scheduled for Anker Day on May 21.

Dubbed the "first Compute-in-Memory (CIM) AI audio chip with neural networks," Thus mimics neurological function by co-locating data storage and processing within its NOR flash memory cells. This architecture avoids the separation typical in standard chips, offering faster read speeds and requiring minimal physical space, perfect for tight component layouts in audio gear.

Headphones present an extreme test case for this technology, demanding consistent performance like active noise cancellation within milliwatt power budgets. Anker specifically announced a feature called Clear Calls, which uses a large, on-device neural network, eight microphones, and two bone conduction sensors to drastically clean up voice quality in noisy settings.

Successfully implementing this level of on-device AI in audio products would serve as a powerful proof point for the Thus architecture. Anker plans to leverage this technology across its broader portfolio of mobile accessories and Internet of Things devices going forward.