GPU Cluster for Simulations and Machine-learning
Abstract
Funding for a GPU cluster is requested to support ongoing DoD related research and student research education (including new academic initiatives in computational physics/chemistry/materials science/biology), in areas currently supported by DoD using state-of-the-art computing algorithms for the study of emerging 2D and interfaces materials. This effort is spearheaded by Clark Atlanta UniversityÕs (CAUÕs) Center for Functional Nanoscale Materials (CFNM), a national Center of Research Excellence for Science and Technology (CREST). CFNM researchers make important contributions in the areas of nanotechnology, and novel theoretical methods in materials theory. Areas being developed include complexity theory analysis, machine-learning algorithms, theoretical/computational support for experimental programs in interface physics and computational biology. The computational study will be crucial for supporting the ongoing research project in multiscale modeling of nanomaterials and theoretical biology. The architecture of General-Purpose Graphics Processing Units (GPGPUs) allows for a significant increase in the performance of numerical computations. In recent years, these systems have gained wide popularity in the science community for supercomputing. GPU computing becomes crucial for the machine-learning algorithm. The requested system will provide the platform to develop and evaluate GPU specific algorithms concerning the research areas. It will also significantly increase CAUÕs computational resources and enable computations using Materials-by-Design methodology.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 04, 2019
- Source ID
- W911NF1910502
Entities
People
- Xiao-qian Wang
Organizations
- Army Contracting Command
- Clark Atlanta University
- Office of the Secretary of Defense