Location: Office of Associate Administrator
Project Number: 0500-00096-001-005-A
Project Type: Cooperative Agreement
Start Date: Sep 29, 2023
End Date: Sep 28, 2027
Agriculture research is rapidly expanding in complexity and scope, especially in the area of agricultural productivity, where it is now possible to include a wide range of variables and massive datasets in relevant scientific research. Historically, productivity researchers largely focused on genotype and environment interactions, but new technology is allowing for the addition of new variables, such as management practices. This calls for the generation of new datasets via a variety of technologies and methods (e.g., unmanned aircraft systems (UAS)/unmanned aerial vehicles (UAV), microbiome sequencing, crop genetics, climate data, animal waste and other items related to crop productivity), the development of new analysis techniques (e.g., artificial intelligence [AI]/machine learning, geospatial statistics, bioinformatics) and the need for high-performance computational systems. To fully realize this new potential to advance agricultural research, complex and massive data sets are needed. In turn, analyzing such data requires access to high-performance computing (HPC) for complex data analysis, modeling, and AI/machine learning. Also, computational tools need to run optimally in an HPC environment, and this takes specialized expertise and development. Furthermore, all of these analytical capabilities require training for proper analysis. A specific example of these needs is the development and use of new statistical and image analysis tools for georeferenced agricultural data related to agricultural operations and productivity.
USDA-ARS has extensive experience researching agriculture productivity (especially with its Long-Term Agroecosystem Research network), microbiome research related to animal waste and field studies, crop genetics, unmanned aerial vehicles, and complex agricultural systems. Mississippi State University (MSU) has expertise in agriculture research and computational methods for agricultural research, especially in the areas of bioinformatics and geospatial statistics. MSU also has extensive experience with the establishment and operations of high-performance computing (HPC) and developing/altering computational software to run in an HPC environment. In this project, MSU will, in consultation with ARS, develop and implement HPC infrastructure that will support the analysis of big data generated by ARS to improve agriculture productivity. Types of analysis include geospatial statistics, bioinformatics, modeling, and AI/machine learning. These types of analysis require specialized HPC infrastructure, including high-memory nodes, normal nodes, graphics processing unit (GPU) nodes, and high-speed data storage.