From the infinitesimal to the galactic: Mira gives scientists a new window to the world

By Laura WolfNovember 15, 2011

ARGONNE, Ill. —The use of supercomputers to propel innovation in science and engineering is an endeavor punctuated by major transformative technologies—the latest being the new open science petascale supercomputers coming online within the U.S. Department of Energy national laboratory complex. These machines will push scientific discovery to a new realm.

Argonne National Laboratory's IBM Blue Gene/Q supercomputer, Mira, is an engineering marvel whose unique architecture and capabilities will be thoroughly explored as soon as it goes online in 2012. Supported by the Department's Office of Science, it will rank among the world's fastest and most energy-efficient supercomputers and represents a milestone in the effort to develop exascale systems equipped with hundreds of millions of processors within the decade.

Access to Mira and the other DOE leadership facilities—awarded in multi-million-hour allocations through the Office of Science-supported INCITE program—is sought after by scientists worldwide to carry out highly complex simulations. Mira, a 10-petaflops supercomputer, will be 20 times faster than Argonne's current leadership-class supercomputer.

To ensure that science applications will be prepared to run as soon as Mira is commissioned, Argonne Leadership Computing Facility (ALCF) staff is working with 16 research teams from across the nation to port and tune their codes on Blue Gene/Q prototype hardware. These Early Science Program projects cover a range of scientific fields representative of Mira's projected computational workload, including simulations of advanced materials, exploration of the universe, modeling of biological organisms, and the design of new, safe, and reliable sources of energy.

  • Argonne materials scientist Larry Curtiss is running electronic structure simulations—calculations capable of predicting the fundamental properties of matter—to identify new materials for electrical energy storage and catalysis, a key component of energy technologies such as biomass conversion and chemical synthesis.
  • Another Early Science project simulates turbulent airflow across aircraft and wind turbine blades. University of Colorado-Boulder's Kenneth Jansen is testing small attachable devices called synthetic actuators that can tweak and control the cross-surface airflow to increase efficiency and reduce damage.
  • University of Chicago astrophysicist Donald Lamb is running simulations of turbulent nuclear combustion—a physical process that is critical to thermonuclear-powered (Type Ia) supernovae. In current models, turbulent nuclear combustion sets the stage for a detonation that incinerates the star, causing it to explode. The simulations use FLASH, a specialized computer code that is able to treat the complex physical processes that occur in these explosions.
  • In the biological domain, University of Chicago and Argonne scientist Gregory Voth conducts multiscale simulations of biostructures, building up from molecular dynamics. His latest research complements the experimental studies of microtubules performed at Argonne's Advanced Photon Source Voth's ultimate goal is to model the biophysics of processes acting at the scale of living cells.

"The visionary research goals of the computational scientists and engineers who gain access to Mira—coupled with the expertise to take advantage of its architecture and capabilities—will enable us to fulfill our mission: to accelerate major scientific discoveries and engineering breakthroughs for humanity by designing and providing world-leading computing facilities in partnership with the computational science community," said Paul Messina, director of science at the ALCF.

The Blue Gene program is a long-term collaborative effort of IBM Corporation, Argonne, and Lawrence Livermore National Laboratory. Blue Gene's speed and expandability have enabled industry and the scientific community to address a wide range of complex problems and make more informed decisions.