Register for free to
view the full innovation profile.
The analog memory element in the circuit stores the resistance state unless changed by application of external voltage or current, greater than a threshold.
About
Dynamic Synapse Circuits U.S. Patent Pending 61/973,754 SU File Reference #152 Abstract Boise State University has developed a Complementary Metal Oxide (CMOS) circuit that solves a fundamental problem in practical development of Neuromorphic or brain-like computing. This circuit uses a compact synapse, which can be adapted to emulate several types of biological synapses and their receptors in human brain. This circuit immediately solves several architectural problems in Machine Learning hardware. Further, the synapse circuit can operate at very high speeds (up to several GHz’s), which is essential for several applications involving Big Data analytics. Applying DC or pulsed voltage or current can alter the resistance of the synapse across its terminals. The analog memory element in the circuit stores the resistance state unless changed by application of external voltage or current, greater than a threshold. The circuit also acts as a compact emulator for the fourth fundamental circuit element called a ‘Memristor’ and resistive memory devices (ReRAMs); and implements the spike-based learning and weight storage functionality of the biological synapses found in an animal brain. The circuit can also synergistically work with memristor devices for fast data processing and non-volatile storage of learned weights in a non-volatile device such as a conducting bridge type memristors, ReRAM, Flash memory, or phase change memory. The dynamic synapse circuit realizes a two-terminal variable resistor, with incremental memory in a feedback loop, whose resistance is controlled by external electrical stimulus applied across the two terminals. The synapse circuit can be designed and fabricated using standard commercial CMOS technologies and used for realizing (i) chip-scale circuits for implementing machine learning algorithms, and (ii) Neurobiology-inspired electronic circuits which can emulate cognitive computing functionality of a biological brain. Since the circuit is completely designed using existing CMOS technology, no additional fabrication steps are needed to design larger neural inspired computing chips. This makes it suitable to essential prototype applications targeted for the conceptual memristor device/element. Boise State is looking for a Licensee for this technology.