The Defense Advanced Research Projects Agency has asked industry for proposals to help DARPA in its push to move forward scientific simulations of physical systems in an effort to carry out complex designs.
DARPA said Friday the Accelerated Computation for Efficient Scientific Simulation program’s goal is to develop a new processor with updated simulation capacity for design, prediction, and discovery for use in partial differential equations that describe fundamental physical principles such as motion, diffusion, and equilibrium.
“Supercomputers today face bottlenecks in converting physical systems into and out of binary form,” said Vincent Tang, DARPA program manager.
“We are going to explore if there are fundamentally better ways to solve multi-scale partial differential equations that describe complex physical systems, such as those encountered in plasmas and fluid dynamics.”
The ACCESS program seeks new ideas for computational architectures classified in traditional architectures as “strong” scaling for predictive scientific simulations, DARPA said.