news, events, and random announcements
Postdoctoral Position Available
The Medford Group currently has an opening for a postdoctoral position at Georgia Tech. The project involves training of machine-learning force fields for prediction of iron surface energies relevant for stainless steel corrosion based on density functional theory calculations. The position also involves efforts toward the development of a new DFT finite difference code with built-in machine learning capabilities for exascale computing. The initial appointment is one year with opportunities to extend based on funding and performance. The position includes competitive compensation and benefits provided through the University System of Georgia. Details of the position:
Tasks to be completed:
Density functional theory calculations of iron surfaces using standard codes
Training of machine-learning force fields based on the AMP package
Extension of AMP to include more efficient training procedures and machine-learning models
Molecular dynamics simulations using AMP-derived force-fields
Development of novel methods for using machine-learning to accelerate DFT calculations
Integration of machine-learning methods with SPARC DFT code that is currently under development
Ph.D in any scientific or engineering discipline
Significant experience in Python programming language and scientific software development
Knowledge of electronic structure theory and calculations
Experience in machine-learning as applied to chemical or atomistic systems
Basic experience in molecular dynamics simulations
Knowledge of surface chemistry
Applications should be addressed to email@example.com. Applications should include a recent CV as well as a link to a Github profile or other evidence of software development experience as appropriate.
The Medford group will collaborate on two new DOE Computational Chemical Sciences projects
The Medford group was part of two successful teams in the recent DOE CCS awards to develop software for exascale computing.
The project “SPARC-X: Quantum simulations at extreme scale —reactive dynamics from first principles” led by Prof. Phanish Suryanarayana at Georgia Tech will develop a new finite-difference DFT code called “SPARC-X” with integrated machine-learning acceleration and optimizations for exascale computing.
The project “Bridging the time scale in exascale computing of chemical systems”, led by Prof. Andrew Peterson at Brown University will establish a new framework for adaptive machine-learning of DFT force fields to tackle the problem of electrochemical reactions occuring at the electrified solid-liquid interface.
Workshop on "Transformative approaches for distributed nitrogen fixation & fertilizer production"
This workshop, organized by Prof. Medford in collaboration with the International Fertilizer Development Center (IFDC) focused on potentially transformative approaches to harness electrical, solar, and mechanical energy to transform air into fertilizers in a distributed way. The workshop took place on Feb. 9, 2017 at the IFDC headquarters in Muscle Shoals, Alabama and included researchers from IFDC, Georgia Tech, and Oak Ridge National Laboratory.
Attendees of the nitrogen fixation workshop. From left to right: Job Fugice, Christian Dimkpa, Porfirio Fuentes, Carsten Sievers, Joaquin Sanabria, AJ Medford, Marta Hatzell, Upendra Singh, Adam Rondinone, Ramon Lazo de la Vega, and Sampson Agyin-Birikorang,