At Nearly 10 Petaflops, Stampede Provides Advanced Computing for US Scientists
AUSTIN, Texas - A National Science Foundation (NSF)-supported world-class supercomputer named Stampede, which has already enabled research teams to predict where and when earthquakes may strike, how much sea levels could rise, and how fast brain tumors grow--was officially dedicated on March 27, 2013. The ceremony, held at The University of Texas at Austin's Texas Advanced Computing Center (TACC) and attended by corporate, government and university leaders, formally introduced Stampede to the public.
The Stampede system has been operational since January 7 with more than 600 active scientific and engineering projects — and well over 1,000 researchers — already using the system to execute simulation and data analysis applications to make new discoveries.
Stampede, a massive Dell/Intel cluster, is a centerpiece of NSF's investment in an integrated advanced cyberinfrastructure, which empowers America's scientists and engineers to interactively share advanced computational resources, data and expertise to further research across scientific disciplines. Stampede is now the most powerful and capable of the 16 high performance computing, visualization and data analysis resources in the NSF-funded Extreme Science and Engineering Discovery Environment (XSEDE) partnership.
"Cyberinfrastructure has increasingly become a critical component of the science and engineering enterprise and is essential to accelerating the pace of discovery and innovation in all fields of inquiry," said Farnam Jahanian, head of NSF's Directorate for Computer and Information Science and Engineering.
"Stampede is an important part of NSF's portfolio for advanced computing infrastructure enabling cutting-edge foundational research for computational and data-intensive science and engineering. Society's ability to address today's global challenges depends on advancing cyberinfrastructure."
Joining Jahanian in dedicating Stampede were U.S. House Science, Space, and Technology Committee Chairman Lamar Smith; Dell Enterprise Solutions President Marius Haas; Intel Senior Vice President Diane Bryant; UT Austin President Bill Powers; and TACC Director Jay Boisseau, who is also the lead principal investigator (PI) of the Stampede project.
"Stampede has been designed to support a large, diverse research community," Boisseau said. "We're as excited about Stampede's comprehensive capabilities and its high usability as we are of its tremendous performance. Stampede will lead the way to major advances in all fields of science and engineering. It's an honor to be at this intersection of advanced computing technologies and world-class science, and we thank NSF, Dell, and Intel for their roles in helping TACC design, deploy, and operate Stampede."
The other PIs from TACC on the Stampede project are Dan Stanzione, Tommy Minyard, Karl Schulz and Bill Barth.
"Stampede is the fastest machine we've ever seen," said Klaus Schulten, a leading expert in physics and director of the Center for Macromolecular Modeling and Bioinformatics at the Beckman Institute for Advanced Science and Technology, and co-director of the Center for the Physics of Living Cells at the University of Illinois at Urbana-Champaign.
"During the early user period, we performed molecular dynamics simulations on biomolecules as a computational microscope. We aim to not only unravel fundamental problems such as how a ‘new born' protein folds, but also solving today's in pressing questions like designing enzymes to produce second-generation biofuel."
Stampede expands the variety of data-intensive, computationally-challenging science and engineering applications that can be used with current national resources. It accommodates large-scale simulations that produce more accurate results.
Stampede has a theoretical peak performance of nearly 10 petaflops, enabled by utilizing two different Intel processor technologies in Dell's new PowerEdge C8220 servers. The base cluster comprises 6,400 nodes with two Intel Xeon E5 processors each, providing 2.2 petaflops of computing power. It also offers powerful remote visualization capabilities, large shared memory nodes, and a large, comprehensive set of software applications and tools. This base system has already been accepted by NSF, and has successfully executed more than 450,000 computational jobs to date.
In addition, Stampede adds a new, innovative technology: the highly parallel Intel Xeon Phi coprocessors. The 6,880 Xeon Phi coprocessors, which are in user evaluation mode now, add more than seven additional petaflops of performance to Stampede, making it the largest installation of Intel's new Xeon Phi highly parallel coprocessors in the world. It is also the largest Dell public production cluster deployed to date. Stampede is the fifth world-class system deployed by TACC, Dell and Intel and supported by NSF to enable open science since 2003.
"We at NSF are gratified to fund such a powerful combination of systems and services for open science research," said NSF Program Manager Irene Qualters. "The technological capacity is important, but even more important is that the scientific community—those on the frontlines of cutting edge, multidisciplinary, science and engineering research to address society's greatest challenges--have open access in order to push the frontiers of science and engineering."
Researchers from any U.S. open science institution can apply to use Stampede. The system will support more than 1,000 projects annually in computational and data-driven science and engineering.
Following are some examples of the exciting and promising early research on Stampede:
Seismic Hazard Mapping - Researchers from the Southern California Earthquake Center (SCEC) are using Stampede to predict the frequency of damaging earthquakes in California for the latest Uniform California Earthquake Rupture Forecast (UCERT3). "We do a lot of HPC calculations, but it's rare that any of them have this level of potential impact," said SCEC's Thomas Jordan. UCERT3 forecasts the probability of all earthquakes throughout a region and over a specified time span for incorporation into the U.S. Geological Survey's National Seismic Hazard Maps which are used to set building codes and insurance rates.
Ice Sheet Modeling to study climate change - The melting of the ice sheets of Antarctica has the potential to raise sea-levels significantly. Omar Ghattas and his team at the University of Texas at Austin are using Stampede to better understand and represent the flow of ice from Antarctica using numerical models.
Improving the Imaging Quality of Brain Tumors - Surgeons want to know how aggressive a tumor is, and the degree of tissue infiltration surrounding the tumor, to be able to plan for surgery, radiotherapy and other treatment options. George Biros at UT Austin is using Stampede to improve the quality of brain tumor imaging so surgeons can make better-informed decisions about treatment options.
Carbon Dioxide Capture and Conversion - The Massachusetts Institute of Technology's Alexie Kolpak is using Stampede to design a new class of nanomaterials and to observe how they perform in real world applications, including solving energy problems.