The Energy Department‘s Argonne Leadership Computing Facility has launched a call for proposals on the development of data science and machine learning applications to support an exascale supercomputer machine.
ALCF said Thursday it will accept proposals for projects related to data science, machine learning and crosscutting simulation data and learning through April 8 as part of the facility’s Aurora Early Science Program.
Aurora ESP will work to prepare the architecture and scale applications of ALCF’s Aurora supercomputer as well as to solidify libraries and infrastructure that will support the system.
The facility kicked off the program last year with 10 simulation-based projects.
ALCF seeks an additional 10 projects in the areas of data analytics, data-intensive computing, advanced statistical analysis, machine learning, deep learning and neural networks.
An Intel–Cray team is building the Aurora machine slated for delivery to the Argonne National Laboratory in 2021.
The system is designed to store more than 5 petabytes of data and perform a quintillion calculations per second.