Magnolia Electronics is a US-based startup developing Analog to Digital converter technology that approaches the conversion task in a new way. We split apart analog signals into simple features, then use machine learning to reassemble those features into the digital values of the analog input. By using lots of simple parts together, our devices work faster and better than traditional devices.
Our non-provisional patent applications covering circuit architectures and training methods are now available from the USPTO.
These documents set out how to build the core parts of our technology, and set out some of the exciting possibilities of what our system can do.