High Performance Computing

The HEP-CCE is highly involved with ASCR HPC computing sites like the at Argonne Leadership Computing Facilities (ALCF) and Oak Ridge Leadership Computing Facilities (OLCF) and the National Energy Research Scientific Computing (NERSC) center at BerkeleyLab.

ASCR/HEP Exascale Requirements Review Report

The HEP-CCE recently participated in the Exascale Requirements Review in conjunction with ASCR. This effort is to help craft the roadmap of the DOE supercomputing facilities for the next ten years in preparation for the Exascale resources that will be deployed. Each science domain was requested to partitipate.


The HEP-CCE has many partners with experience using supercomputers for physics simulation and analysis. This includes successfully competing and being awarded computing time on DOE computing resources.

  • Salman Habib and Katrin Heitmann who are leading a group at Argonne National Laboratory to perform Cosmic Frontier simulations at some of the largest scales at many of the DOE ASCR scientific computing facilities. [article,article]
  • Tom LeCompte and Taylor Childers are leading a group at Argonne to run LHC collision event generation and simulation on the Mira and Edison supercomputers. [article]
  • James Amundson is leading a team at Fermilab to run accelerator simulations on the Mira supercomputer. [article]

Facilities Available

Argonne Leadership Computing Facility hosts the fifth fastest supercomputer. Mira is a IBM BlueGene/Q with 768k PowerPC cores and high speed interconnects. The Theta test system, the test system for Mira’s replacement, Aurora, will be deployed in Fall-Winter 2016.

NERSC at Berkeley hosts the Edison and Cori supercomputers.

Oak Ridge Leadership Computing Facility hosts the second fastest supercomputer in the world. Titan is a hybrid of CPUs and GPUs containing  18,688 AMD Opteron CPUs (16-cores each) each paired with one NVIDIA Tesla K20X GPU.

See Training Page for how to get started

An HEP Collision Point