RT service highlight: Big Red 200

With artificial intelligence capabilities and a peak performance rate of more than 6 petaFLOPS, IU’s latest supercomputer solidifies its standing as a high performance computing powerhouse.

Use case: IU unveils supercomputer Big Red 200

Big Red 200 will operate at a peak rate more than six times faster than its predecessor (Big Red II), with greater than 6 quadrillion—or 6 thousand trillion—floating-point operations per second, or petaFLOPS. Named Big Red 200 in honor of the IU Bicentennial celebration, the new system is more than 300 times faster than the original Big Red supercomputer installed 15 years ago. IU dedicated Big Red 200 as part of IU’s Bicentennial event, “A Day of Commemoration: IU’s 200th Anniversary,” on January 20, 2020.

RT service highlight: Big Red II and Big Red 3

2019 saw the retirement of Big Red II, once IU’s primary system for high performance parallel computing dedicated to research. In five years, it supported over 1,700 users in 237 academic disciplines, handling more than 2.5 million jobs with over a billion core hours delivered. Its successor, Big Red 3, is a Cray XC40 supercomputer dedicated to researchers, scholars, and artists with large-scale, compute-intensive applications that can take advantage of the system’s extreme processing capability and high-bandwidth network topology.

• 930 dual-socket compute nodes equipped with Intel Haswell Xeon processors (22,464 compute cores)
• Theoretical peak performance (Rpeak) of 934 trillion floating-point operations per second (934 teraFLOPS)

Use case: Physicists explore the nature of quantum physics with Big Red II and 3

At IU, physicists are taking a closer look at how electrons behave, with help from hundreds of millions of calculations powered by HPC clusters Big Red II and 3. Their work expands upon recent discoveries that proved the existence of the Hofstadter Butterfly, a fractal pattern that shows the behavior of electrons in a magnetic field. IU physicists are hoping to classify the types of electrons present in a magnetic field.

“In these types of classifications, the discovery is made possible by supercomputers like Big Red II and 3. The intricacies of what is happening in a physical system are too great for us to have any equation we can solve on paper. We need to use the computers to solve and find these values.”

Babak Seradjeh
Associate Professor of Physics, IU Bloomington

RT service highlight: Carbonate

Carbonate is a large-memory computer cluster configured to support high performance, data-intensive computing. Carbonate can handle computing tasks for researchers using genome assembly software, large-scale phylogenetic software, and other genome analysis applications that require large amounts of computer memory.

• 72 general-purpose nodes, 256 GB RAM each
• 8 large-memory compute nodes, 512 GB RAM each
• Each node: Lenovo NeXtScale nx360 M5 server with two 12-core Intel Xeon E5-2680 v3 CPUs and four 480GB SSDs

Use case: Exploring gene expression through high performance computing

R. Taylor Raborn and his colleagues design methods for locating gene promoters. They use Carbonate to develop containerized deployment and application of their analysis software.

“This wouldn’t have been possible without the great resources that UITS provided for us. I don’t know if there’s any high performance computing center in the country that is as easy to work with and that does as much for the people that work in the university as UITS….It really democratizes the use of resources.”

R. Taylor Raborn
Research Scientist, Arizona State University Biodesign Institute

Use case: IU researchers study role of microbiomes in honeybee function

Eric Smith, a postdoctoral fellow in the IU biology department’s Newton Lab, studies how honeybees’ microbiomes support and influence their function. Smith explores the specifics of these symbiotic relationships using computational biological methods. He uses IU’s large-memory computer cluster, Carbonate, to write and run computational programs to analyze DNA sequences.

“Carbonate provides the computational infrastructure necessary to perform large amounts of data simulation to test computational pipelines and eventually analyze DNA sequencing data.”

Eric Smith
Postdoctoral Fellow in Biology, IU Bloomington

RT service highlight: Carbonate deep-learning nodes

Starting in June 2019, the Carbonate Deep Learning resource delivered 759,688 core hours and 92,783 GPU hours to users conducting research in disciplines including medical image segmentation, video classification, cybersecurity, genomics, and natural language processing.

“In its first year, the Deep Learning expansion of the Carbonate system enabled research in over 130 projects. With its uniquely capable V100 GPUs, this resource gives IU’s researchers the ability to get ahead of the curve with their research using AI techniques.”

Scott Michael
Manager, Research Applications and Deep Learning, IU Research Technologies

Use case: Identifying neutrinos with machine learning

Mark Messier, Micah Groh, and Ryan Murphy of the IU physics department measure the rates at which neutrinos change their type, or flavor. By looking at the particles’ patterns, they hope to determine if matter and antimatter neutrinos follow the same pattern, or if they differ from each other. The team uses Carbonate to process examples of neutrino interactions, feeding them through an algorithm so that it can learn each flavor’s identifying characteristics.

“The Carbonate computer cluster here at IU has top-of-the-line GPU processors, which are perfect for processing these examples quickly. Having this resource readily available at IU allowed us to try lots of ideas out so we could find the best ones.”

Micah Groh
Graduate student in Physics, IU Bloomington