atomically Thin nano-scale resistive memory
Resistive memories are two terminal nano-devices which have similar characteristics as biological neurons and they are building blocks of our brain-inspired AI processors. Our devices are fabricated using novel 2-dimensional materials which gives higher energy-efficiency and reliable performance.
highly-connected memristive networks
Processor core is network of highly connected memristive sub-cores where each sub-core focuses on analyzing different type of signal. This structure emulates the brain processing structure and hence makes our processors 100x energy-efficient and powerful comparing to current processors.
novel processing architecture
Our processors have hybrid implementation of deep neural networks, spiking neural networks, and reinforcement learning that gives adaptable learning capability directly on the hardware. So, our flexible AI processor improves safety of autonomous vehicles, makes service/industrial robots more robust, and gives efficient flight operation to drones.