DARPA is seeking innovative research proposals in the area of ultralow power in-memory computing architectures for application to scientific modeling, simulation, and analysis. The NaPSAC program aims to develop novel in-memory computing architectures capable of transformative advances in computing accuracy, scalability, and power efficiency.
It is anticipated that performers in this program will validate and benchmark compact, efficient in-memory computing engines capable of performance beyond the conventional state-of-the-art (SoA) von Neumann approaches in scientific modeling of complex, multi-spatiotemporal, and nonlinear problems. Examples of such problems include advanced materials discovery, chemical synthesis, radiation electrodynamics, fluid dynamics, energy or mass transport through heterogeneous media, semiconductor device design and fabrication, and earth-system models.
This is a two-phase program, with each phase expected to run for 24 months.
– Abstracts due 16 February 2023, full proposals due 4 April 2023