Managing computational materials science in the high-throughput era: AiiDA and the Materials Cloud

Data: 
26/03/2018 - 14:00 até 15:00
Palestrante: 
Dr. Spyros Zoupanos, Ecole polytechnique fédérale de Lausanne, Switzerland
Local: 
Sala de Seminários José Roberto Leite - Edifício Alessandro Volta (bloco C)
Resumo: 

In recent years, there has been a great increase in the performance of computers and the development of accurate simulation codes. Materials science has greatly benefited from these advances, which are boosting research and the discovery of new materials. The "materials by design" approach has become very powerful, but requires running large numbers of simulations and building databases of computed properties. A key challenge is the need to automatically prepare, execute and monitor workflows of calculations, and then retrieve and store the results in a format that is easy to browse and query. Such infrastructure should offer a fast learning curve and encourage sharing of codes, results and workflows. In this talk I will illustrate the platform requirements, that we have identified in the four "ADES" pillars: Automation, Data, Environment, and Sharing. I will then describe how we have implemented these in our open-source AiiDA platform [1] whose design, based on directed gra!

 phs, aims at storing the complete provenance of the research data, facilitating its analysis to accelerate scientific discovery. I will briefly present the Materials Cloud [2], a web platform designed to enable seamless sharing of resources in computational materials science. This includes educational materials like videos and lectures (Learn section); interactive tools and virtual hardware to run and manage simulations in the cloud or offline (Work section); access to curated data (Discover section) together with full access to their provenance as tracked by AiiDA (Explore section); and long-term storage or research data with DOIs assigned to them (Archive section).

 

Desenvolvido por IFUSP