Exploration flight projects, robotic precursors, and technology demonstrators that are designed to operate beyond low-earth orbit (LEO) require avionic systems, components, and controllers that are capable of enduring the extreme temperature and radiation environments of deep space, the lunar surface, and the Martian surface. This SBIR effort will provide a low-cost, radiation-hardened, non-volatile memory technology tolerant to extreme temperature ranges for all NASA space missions that require storage and processing of large amounts of data. The proposed innovation addresses the NASA technology needs outlined in OCT Technology Area TA11: Modeling, Simulation, Information Technology and Processing Roadmap, in particular, for Computing (Flight Computing, high performance space-based computing), which requires ultra-reliable, radiation-hardened platforms which have been costly and limited in performance. Other important products of immediate impact to NASA include: Computer Aided Design (CAD) tools for predicting the electrical performance of low-temperature and wide-temperature electronic components and systems; and physics-based device models valid at temperatures ranging from -230 deg C to +130 deg C to enable design and verification of robust radiation-hardened memory circuits.
This project will enable significant progress towards the use of memristor-based systems in a wide range of non-NASA aerospace and defense applications that require storage and processing of large amounts of data. The critical question about the combined radiation and temperature tolerance of different memristor technologies will be answered, paving the way for the development of memristor-based non-volatile memory, threshold logic, and reconfigurable architectures (FPGAs) for space applications, such as broadband communication, surveillance, image processing, etc. The improved, physics-based modeling and simulation tools, applicable to both chalcogenide-based and transition-metal-oxide based memristor technologies, will allow designers to perform fast, reliable, and more accurate characterization of memristor-based circuits as a function of various stress conditions (i.e., bias, thermal, and radiation). The generated compact models will also be a significant aid for circuit design/analysis. The simulation and design tools will benefit manufacturers of commercial satellite electronics and avionics, where the memristor is a strong candidate for static RAM, as it combines the advantages of the hard disk (density), RAM (access speed), and flash-based memories (low power, non-volatile).