Autonomous Computing Materials
Mark Bathe,Rigoberto Hernandez,Takaki Komiyama,Raghu Machiraju,Sanghamitra Neogi
DOI: https://doi.org/10.1021/acsnano.0c09556
IF: 17.1
2021-03-23
ACS Nano
Abstract:Conventional materials are reaching their limits in computation, sensing, and data storage capabilities, ushered in by the end of Moore's law, myriad sensing applications, and the continuing exponential rise in worldwide data storage demand. Conventional materials are also limited by the controlled environments in which they must operate, their high energy consumption, and their limited capacity to perform simultaneous, integrated sensing, computation, and data storage and retrieval. In contrast, the human brain is capable of multimodal sensing, complex computation, and both short- and long-term data storage simultaneously, with near instantaneous rate of recall, seamless integration, and minimal energy consumption. Motivated by the brain and the need for revolutionary new computing materials, we recently proposed the data-driven materials discovery framework, autonomous computing materials. This framework aims to mimic the brain's capabilities for integrated sensing, computation, and data storage by programming excitonic, phononic, photonic, and dynamic structural nanoscale materials, without attempting to mimic the unknown implementational details of the brain. If realized, such materials would offer transformative opportunities for distributed, multimodal sensing, computation, and data storage in an integrated manner in biological and other nonconventional environments, including interfacing with biological sensors and computers such as the brain itself.