Publications

Chris Saunders and three technologists are in high demand from Sandia’s deep learning teams, and they’re kept busy by building new clusters of computer nodes for researchers who need the power of supercomputing on a smaller scale. Sandia researchers working on Laboratory Directed Research & Development (LDRD) projects, or innovative ideas for solutions on short timeframes, formulate new ideas on old themes and frequently rely on smaller cluster machines to help solve problems before introducing their code to larger HPC resources. These research teams need an agile hardware and software environment where nascent ideas can be tested and cultivated on a smaller scale.

Sandia has been developing and supporting data transfer tools for over 20 years and has the expertise to take DOE into the Extreme Scale era. In looking at Exascale and beyond (Extreme Scale Computing), data sets can be thousands of 500TBs in size, a single file can be in the 100TB range, and billions of files are expected. Huge bursts of data need to be transferred, even today. While data archiving is often not thought about, it is an integral part of the full data management path when data is generated on HPC systems. In order to move generated data to its final resting place (data archive) or to transfer between file systems, a capable data transfer tool is required.

Carnac, located at Sandia's California site, is an institutional cluster for Emulytics that provides security researchers with resources to model enterprise computer networks and evaluate how resilient they are from attacks. While multiple Emulytics cluster computers have been built at Sandia, Carnac is the first system that was developed as an institutional resource that can be shared among different groups with disparate requirements.

A patient in the United States has been diagnosed with Ebola. Fear and panic kicks in across the country, and hospitals are inundated with hundreds of people, some infected with the highly contagious disease and others not. Blood tests are needed for positive diagnoses, but the diagnostic labs are overwhelmed with blood samples to test, and staff are overworked and stressed. Infected people need to be quarantined and treated, but it's hard to find rooms to quarantine so many patients. Sick people who need triage and regular care for other emergencies are afraid to go to hospitals for fear of Ebola, which has a 50% fatality rate. And since hospitals are so overwhelmed, sick people often stay home, infecting heathy people around them; the U.S. is now in the grips a full-blown Ebola outbreak. Sandia's high-performance computers simulated such a nightmare scenario recently, and with good reason. An Ebola outbreak in the United States could be devastating if hospitals are not prepared. When an Ebola outbreak in West Africa became a global concern in 2014, health advisers were alarmed at the length of time it took to properly diagnose infected people. In rural areas in Liberia, for example, blood samples from ailing people would be sent to a laboratory for testing, but the closest lab was hundreds of miles away through difficult and sometimes impassable roads. In more urban areas, blood samples would be sent to nearby labs, but those labs were often already overburdened by the sheer volume of samples to test. Staff at some treatment centers were unaware that a lab a little farther away might have the capacity to take in more samples. Meanwhile, undiagnosed infected people were unknowingly spreading the disease to many others around them, worsening the outbreak. The U.S. Defense Threat Reduction Agency (DTRA) and Centers for Disease Control and Prevention (CDC) posed a serious question: how do we improve blood-sample transportation routes in Liberia to ensure that samples taken from ill people are tested as quickly as possible, ensuring a proper diagnosis and faster treatment? Sandia scientists, already experts in transportation modeling for nuclear materials, quickly swarmed on this problem. The Sandia Ebola response team immediately set out to collect data from the region using available maps and local information, and transformed the raw data to GIS maps. Then, applying Sandia transportation routing algorithms, the team identified the optimal routes to get blood samples to the best laboratory for testing, even if that lab was not geographically the closest. The models also showed the best possible locations for mobile diagnostic laboratories that would better support the very rural regions that were most affected by the Ebola outbreak.

In October 2017, Sandia broke ground for a new computing center dedicated to High Performance Computing. The east expansion of Building 725 was entirely conceived of, designed, and built in less than 18 months and is a certified LEED Gold design building, the first of its kind for a data center in the State of New Mexico. This 15,000 square-foot building, with novel energy and water-saving technologies, will house Astra, the first in a new generation of Advanced Architecture Prototype Systems to be deployed by the NNSA and the first of many HPC systems in Building 725 East.

Chris Saunders and three technologists are in high demand from Sandia's deep learning teams, and they're kept busy by building new clusters of computer nodes for researchers who need the power of supercomputing on a smaller scale. Sandia researchers working on Laboratory Directed Research & Development (LDRD) projects, or innovative ideas for solutions on short timeframes, formulate new ideas on old themes and frequently rely on smaller cluster machines to help solve problems before introducing their code to larger HPC resources. These research teams need an agile hardware and software environment where nascent ideas can be tested and cultivated on a smaller scale. Saunders and his team at Sandia's Science and Engineering Computing Environments are successfully enabling this research by creating pipelines for emerging code—from Cloud, to containers, to virtual machines—that build the right environment quickly to help teams solve their problems in a matter of days rather than months. While the larger HPC sources are available, it's these smaller clusters that can rapidly build a foundation for teams to build on for later development on larger systems.

A new race is emerging among nuclear powers: the hypersonic weapon. Hypersonics are flight vehicles that travel at Mach 5 (five times the speed of sound) or faster. They can cruise in the atmosphere, unlike traditional exo-atmospheric ballistic missiles, allowing stealth and maneuverability during midflight. Faster, lower, and stealthier means the missiles can better evade adversary defense systems. The U.S. has experimented with hypersonics for years, but current investments by Russia and China into their own offensive hypersonic systems may render U.S. missile defense systems ineffective. For the U.S. to avoid obsolescence in this strategically significant technology arena, hypersonics—combined with autonomy—needs to be a force multiplier. Achieving an autonomous hypersonic missile, however, that can intelligently navigate, guide, and control itself and home-in on targets ranging from traditional stationary systems to targets that are themselves hypersonic vehicles—with all the maneuverability that this entails—may sound far-fetched. But to Sandia's Autonomy for Hypersonics (A4H) team, this dream is one step closer to reality.

The High-Performance Computing of today is much different than the supercomputing resources a few decades ago. The supercomputers of the past were huge complexes of hundreds of interconnected computers that had large teams of specialists to keep them running. Today, while supercomputers like Summit (the world's most powerful supercomputer as of June 2018, installed at Oak Ridge National Laboratory) are still in high demand for extremely fast computational power (over 200,000 trillion calculations per second, or 200 petaflops), researchers at Sandia now have access to non-traditional computing resources in the way of very powerful graphics processing units (GPUs). These GPU systems have enabled an entirely new avenue of exploration for Sandia, allowing researchers to further tackle problems in energy, advanced materials, artificial intelligence, and nuclear weapons. Many teams at Sandia utilize these GPU clusters to build models through the use of machine learning. These models are faster representations of their code-driven counterparts and can often be leveraged from the exascale computing resources of traditionally larger systems with huge boosts in performance. We highlight two resources for these smaller systems: the team that builds the hardware, and the team that researches and builds the machine learning algorithms.

Abstract not provided.

The computational power of HPC is beyond our comprehension when we hear that 5 quadrillion computations can happen in a matter of seconds, or that machine learning is changing the way everything works. But none of that happens in a vacuum, and the teams behind the scenes—the developers of the hardware, the operating systems, the data transfer protocols, and the applications themselves—are the unsung heroes of a world where faster is better and you'd better hope there's no bug in the software or the hardware to slow you down. HPC is most successful when all these aspects work together seamlessly. The stories that follow are a tribute to the hardworking teams behind the scenes.