About

ADC is equipped with two corner AI technologies in two domains: hardware (2-pending patents) and software (3-US patents and 2- pending patents)
Hardware: Hybrid In-memory processing architecture with 8-bit accuracy (On-going provisional patent filling), to enable SWaP-C deep learning processor for cell phone , mobile devices, edge processing and space intelligence. The hybrid high speed correlator is a prime candidate to advance substantially the conventional Von Neumann architecture type through in-memory computing, irrespective of parallelizing efforts (Nvidia), in order to fully exploit the capability of power reduction. (The CTO had 26 years of neural network hardware based on in-memory processing at JPL/NASA from 1985. )
Software: Developing and evaluating optimal machine learning and AI model based on bio-structure model e.g., short-term, long-term memory via feed forward system: unsupervised and supervised learning and feedback system: real-time adaptation and auto associate memory, to enable effective learning in fast conversion, much less computing power and optimal generalization capability for autonomous intelligence with self-training data generation and self-learning capabilities.
The combination of two is expected to put ADC for a true intelligent machine for a coming future.