Dell, University of Texas Benchmark Performance of Standards-Based Supercomputing Clusters

April 11, 2005

Dell and the Texas Advanced Computing Center at The University of Texas at Austin have teamed to benchmark high-performance computing (HPC) applications on standards-based computer hardware. The organizations' goal is to develop a better understanding of the performance and to enhance the scalability of applications run on these cost-effective supercomputers.

"Our relationship with Dell has enabled us to offer tremendous computing capabilities to TACC users in Texas and across the US," said Jay Boisseau, director of TACC. "Our new, cooperative benchmarking project will further enhance the systems we offer to researchers nationwide while assisting Dell in providing the best HPC solutions to the community."

TACC will work with Dell's internal HPC group to enhance TACC's testing of the performance and scalability of various processors and interconnects on HPC applications.

"The knowledge, information, and resource sharing between Dell and the TACC will help us to even more effectively and optimally satisfy the supercomputing needs of the researchers and engineers across higher education, the federal government and the private sector by standard-based computing platforms," said Reza Rooholamini, PhD., director, Enterprise Solutions Engineering, Dell.

While the partnership is initially focused on benchmarking and performance analysis of various hardware technologies, it will eventually expand to include evaluation of HPC software technologies; testing of grid-computing technologies; and options for clusters devoted to visualization and data analysis applications.

"Industry standard-based cluster computing has grown significantly over the past few years because clusters have proven to be very cost-effective solutions for the performance they provide," said Tommy Minyard, Research Associate and Manager of TACC's HPC group. "Researchers can now run their applications at a much larger scale and higher resolution while decreasing total time to solution. With Dell's support, we will continue to investigate and improve the performance of applications run in these clustered computing environments and explore new techniques and algorithms for improving performance."

Last year, TACC deployed Lonestar, one of the world's most powerful computing systems. Lonestar is built with a Dell HPC cluster based on the Intel Xeon processor. TACC recently upgraded the technology powering Lonestar and installed a second Dell cluster, Wrangler, with 64-bit Intel Xeon processors, Myrinet and InfiniBand switches.

Explore further: SDSC's 'Comet' supercomputer enters early operations phase

Related Stories

SDSC's 'Comet' supercomputer enters early operations phase

April 30, 2015

Comet, a new petascale supercomputer designed to transform advanced scientific computing by expanding access and capacity among traditional as well as non-traditional research domains, has transitioned into an early operations ...

Recommended for you

Innovations from the wild world of optics and photonics

August 2, 2015

Traditional computers manipulate electrons to turn our keystrokes and Google searches into meaningful actions. But as components of the computer processor shrink to only a few atoms across, those same electrons become unpredictable ...

Shedding light on millipede evolution

August 2, 2015

As an National Science Foundation (NSF)-funded entomologist, Virginia Tech's Paul Marek has to spend much of his time in the field, hunting for rare and scientifically significant species. He's provided NSF with an inside ...

Better together: graphene-nanotube hybrid switches

August 2, 2015

Graphene has been called a wonder material, capable of performing great and unusual material acrobatics. Boron nitride nanotubes are no slackers in the materials realm either, and can be engineered for physical and biological ...

0 comments

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.