From Texas to Mars, IBM Supercomputer to Power Research

Texas Advanced Computing Center Installing Four IBM eServer p690s and Two IBM Linux Clusters

Select a topic or year

Armonk, NY - 28 Dec 2001: IBM today announced that it will provide the University of Texas at Austin (UT) with powerful new supercomputing systems for scientific research into areas as diverse as manned space flight to Mars and next-generation Internet "Grid" computing applications.

Tripling the aggregate computing power, the Texas Advanced Computing Center's (TACC) systems include an IBM eServer* Cluster 1600 UNIX® system -- expected to be the most powerful academic supercomputer in Texas -- as well as two IBM eServer Cluster 1300 Linux® systems. All of the systems will be available to UT's community of researchers in January 2002.

The POWER 4-based UNIX systems will help the university's Center for Space Research (CSR) prepare the United States for an eventual manned landing on Mars. Scientists at the CSR's Navigation and Control Lab will use the supercomputer to improve the precision of interplanetary navigation, terminal descent and landing on Mars, and modeling of the Martian environment. CSR will use the system to handle the monster number-crunching assignments that are crucial to developing a wide range of techniques needed to extend mankind's ability to guide, navigate, and control spacecraft on both planetary and interplanetary missions.

"The raw performance coupled with the excellent bandwidth of the IBM systems gives our researchers vast improvements in performance," said Dr. Jay Boisseau, director, TACC. "For science and engineering applications, bandwidth is often the key to numerically intensive simulations running more accurately."

Grid Research
The supercomputer systems will also be used to test powerful new computing Grids. Using a variety of Grid computing software, in particular the Globus Toolkit™ for security, resource allocation and data management, one of the Linux clusters will be linked to existing IA32 clusters to test and build a variety of Grid applications.

Grid computing has the potential to fundamentally change computing by enabling heterogeneous systems to share resources over the Internet. In conjunction with the National Partnership for Advanced Computational Infrastructure (NAPCI), TACC's research will supplement the NSF Distributed Terascale Facility (DTF) efforts to accelerate the deployment of Grid computing, nationally and worldwide.

TACC's research is focused on developing mature Grid software to bolster the use of available Grid computing power. Using the IA64 based systems, researchers will run Grid usage models simulating a broad scientific and engineering community and test and develop the software. The results of having stable software for Grid computing will encourage more researchers to embrace Grid computing and its vast potential.

"The scientists at TACC are undertaking important research across a broad spectrum of disciplines," said IBM vice president Peter Ungaro. "The power and reliability of the IBM eServer p690 systems as well as IBM's integrated Linux clusters can greatly accelerate the depth and accuracy of their projects."

The UNIX cluster will be built with four IBM eServer p690 systems, the world's most powerful UNIX servers. Accommodating increased demand from the scientific community, the power of the IBM system will also enhance research and development activities in areas such as oil exploration and space research.
The first of the Linux systems deployed will be an Intel IA64 cluster with 40 Itanium 800 MHz processors with 80GB of memory. The second cluster will be an Intel IA32, based on 64 Pentium III 1 GHz processors with 32 GB of memory.

The IBM eServer Linux clusters will enable researchers in the Texas Institute for Computational and Applied Mathematics (TICAM) to develop more accurate, high-resolution models of underground oil reservoirs. Accurate subsurface models are vital to the nation and the energy industry because they help maximize the recovery of oil while minimizing both the cost and the environmental impact.