Exploring Cloud Paradigm and Practices for Science and Engineering
Manish Parashar, Ivan Rodero, Javier Diaz Montes received a two-year NSF award of $299,984 for the project "Exploring Cloud Paradigm and Practices for Science and Engineering."
The project abstract is given below.
Clouds abstractions and infrastructure are rapidly becoming part of the advanced research cyber-infrastructure (ACI) providing viable platforms for scientific exploration and discovery. As a result, it is important to understand how emerging data and compute intensive application workflows can effectively utilize a hybrid ACI integrating Cloud abstractions and services, and how such a hybrid ACI can enable new paradigms and practices in science and engineering. This EAGER explores innovative science and engineering application
formulations that are enabled by a hybrid federated ACI that includes Clouds and HPC resources, as well as programming and middleware support for these new application formulations. Specifically, we focus on three key research thrusts: (1) application formulation; (2) programming models, abstractions and systems; and (3) middleware stacks and management services, and explore two applications use cases - (i) an oil reservoir modeling application based on an Ensemble Kalman Filter (EnKF), and (ii) molecular dynamics simulations using asynchronous replica exchange. In each of these use cases we explore how the capabilities provided by resources and services in a federated ACI can be leveraged to optimize metrics such as time-to-science, cost-to-science and/or energy-to-science.
Cloud services are integral to the NSF ACI vision. Clouds are also rapidly becoming an integral part of the ACI available to science and engineering applications, and provide complementary capabilities that can have a significant impact on a range of applications. As a result, this research can have a significant impact on a diverse set of application domains by identifying new paradigms and practices that can make effective use of a hybrid ACI to accelerate science. Furthermore, the results of this research will provide resource providers information about how to best meet the needs of science and engineering applications and how current ACI can achieve broader accessibility and higher efficiencies and productivity. The development of human resources, including the training of students, researchers and software professions, as well as outreach to minorities and underrepresented group, is integral to all aspects of this effort.