The new supercomputer system will run computationally intensive climate and biological system simulations, along with other scientific programmes, to aid DOE research in environmental science, atmospheric and climate modelling, bio-energy applications, molecular chemistry and materials science, and more. As a national user facility resource, the system will be available to scientists conducting research that aligns with EMSL's mission to solve the nation's environmental and energy production challenges.
"This flagship high performance computer will prove invaluable in providing innovative solutions to some of the most important national challenges in energy and environmental sciences", stated Gary Skouson, senior research scientist at EMSL. "Having used Moab for our previous supercomputing system, we look forward to employing it again to solve our workload management decisions."
Using its Moab HPC Suite, Adaptive Computing will automate the scheduling, management, monitoring and reporting of workloads on a 1,440 compute node supercomputing system built on Supermicro SuperServers and Intel Xeon Phi accelerators - totalling over 195,000 processors. Each node is capable of delivering processing speeds of 2.3 teraflops.
"Moab provides this environment with workload management software that is efficient, battle-tested and scalable", stated Bart Willems, senior HPC engineer at Atipa Technologies. "This is our first multi-petaflop system containing Intel Xeon Phi accelerators that runs on Moab, and we anticipate it will occupy a top 20 position as one of the world's fastest supercomputers in the next release of the TOP500 list."
Moab HPC Suite is Adaptive Computing's intelligent workload management software, which optimizes scheduling and management across workloads and resources based on policies. Accelerating results delivery and maximizing utilization, Moab HPC Suite simplifies workload management across complex, heterogeneous HPC cluster environments. It leverages multi-dimensional policies to continually schedule and monitor workloads, resources, service-level agreements (SLAs) and priorities to optimize workload output.
"It is exciting that Moab is playing a key role in creating this problem-solving environment for the Department of Energy's EMSL", stated Rob Clyde, CEO of Adaptive Computing. "With the increased efficiency provided by Moab, users can meet diverse job requirements quicker, drive innovation and discovery faster, and further accelerate their research."