Argonne is the central site for work on database and data management. The group has key responsibilities in the design and implementation of the I/O model which must provided distributed access to many petabytes of data for both event reconstruction and physics analysis.
The group deployed a number of HEP packages on the BlueGene/Q supercomputer of the Argonne Leadership Computing Facility, and currently generates CPU-intensive Monte Carlo event samples for ATLAS. Also, the group contributes to the HEP community by creating Monte Carlo samples for the HepSim public repository maintained by the ATLAS group.
The ATLAS group at Argonne collaborates with computer scientists and engineers at Argonne's Leadership Computing Facility (ALCF) which some of the world's most powerful supercomputers. With capabilities many times that of the LHC Computing Grid, we are working to leverage these machines to support ATLAS analyses.
The ATLAS group at Argonne collaborates with computer scientists, mathematicians, and engineers at Argonne's Leadership Computing Facility (ALCF) and Math and Computer Science (MCS) divisions. Our collaborators are interested in tackling challenging science problems that require large scale machine learning on supercomputers and we are interested in benefiting from recent advances applying convolutional neural networks to image recognition problems.