NSF Award
2425538
As society increasingly relies on digital technologies, the growing energy consumption of computing systems makes their continued scaling unsustainable. At the same time, conventional computers face fundamental challenges in solving many important computing problems related to optimization. A possible solution to these challenges is to fundamentally change the architecture of computing systems. A promising example of such unconventional computing approaches is probabilistic (p-) computing, which uses a network of probabilistic bits that collectively evolve towards the network’s energy minima, which are designed to correspond to the solution(s) of the computing problem of interest. However, the realization of large-scale p-computers that provide computational advantage over conventional computers still requires improvements in the energy efficiency and speed of existing p-bits. This project will address this need by developing p-bits based on a new type of magnetic device, referred to as an antiferromagnetic tunnel junction. These devices have inherently faster dynamics than existing magnetic p-bits, making them excellent candidates for p-bit implementation. The project brings together experts in materials, devices, circuits, and architectures, who will co-design the proposed p-bits and explore domain-specific computing architectures that combine these p-bits with state-of-the-art semiconductor chip technology. <br/><br/>The results of this project will impact a broad range of commercial markets, which face hard computational tasks related to combinatorial optimization. Applications of the developed domain-specific probabilistic computers can include logistics, transportation networks, wireless infrastructure, and chip design, to name a few. In addition, this project also contains educational components for semiconductor workforce development. These plans include collaborative development of a new course focusing on next-generation computing based on emerging materials. The project will also collaborate both with external professional societies as well as with local university resources, to provide opportunities for high school, undergraduate, and community college students to gain exposure to scientific research and training in magnetism and advanced computing. Through its combined research and workforce development efforts, the project will contribute to the continued leadership of the United States in the important areas of microelectronics technology, chip design, and advanced computing systems.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
