HPC4Mfg – Computational Design and Optimization of Ultra Low Power Device Architectures

Project Description

Advanced modeling and simulation capabilities and large scale data analysis provide tools to address challenging problems in decision making, process and design optimization, and performance prediction that occur in manufacturing. Effective use of High Performance Computing (HPC) in manufacturing has the potential to improve quality, reduce energy costs, accelerate or eliminate expensive testing, and accelerate the time of adoption of new technologies. With the support of the Department of Energy’s Advanced Manufacturing Office (AMO), Lawrence Livermore National Laboratory (LLNL), Oak Ridge National Laboratory (ORNL,) and Lawrence Berkeley National Laboratory (LBNL) have formed the HPC4Mfg program designed to allow partnerships with industry in bringing HPC into manufacturing.

This award supports the HPC4Mfg program, providing time on HPC resources to three National Laboratory/Industry partnerships designed as pilot programs for applying HPC to advanced manufacturing projects. This specific HPC4Mfg project is described below:

“Computational Design and Optimization of Ultra Low Power Device Architectures,” a partnership between Global Foundries and LBNL, will use atomistic simulation to simulate annealing of amorphous oxide layers, a key step in the production of semiconductors. Improved understanding of this process will enable future electronics with reduced power usage.