Todd Jacobson
NS&D Monitor
11/21/2014
The Department of Energy is investing $425 million in its pursuit of exascale supercomputing, including the award of two new supercomputers at the Oak Ridge and Lawrence Livermore national laboratories. Energy Secretary Ernest Moniz announced the awards late last week, as well as a $100 million boost to DOE’s FastForward 2 program, which shepherds research and development into extreme scale supercomputing technologies. “High-performance computing is an essential component of the science and technology portfolio required to maintain U.S. competitiveness and ensure our economic and national security,” Moniz said in a statement. “DOE and its National Labs have always been at the forefront of HPC and we expect that critical supercomputing investments like CORAL and FastForward 2 will again lead to transformational advancements in basic science, national defense, environmental and energy research that rely on simulations of complex physical systems and analysis of massive amounts of data.”
IBM, NVIDIA, Mellanox to Develop Supercomputers
The supercomputer awards at Livermore and Oak Ridge each will be developed by IBM, NVIDIA, and Mellanox. Livermore’s new supercomputer will be called Sierra and is expected to be seven times more powerful than Livermore’s Sequoia supercomputer. Oak Ridge’s supercomputer will be named Summit and is estimated to be at least five times faster than the lab’s Titan supercomputer, which is the fastest supercomputer in the nation and the second fastest supercomputer in the world, according to twice-yearly rankings released this week. China’s Tianhe-2 supercomputer tops the list of the world’s fastest supercomputers for the fourth time in a row, with Titan coming in second with the capability of more than 20 quadrillion operations per second.
A third supercomputer award will go to Argonne National Laboratory at a later date as part of a joint program with Livermore and Oak Ridge, DOE said. FastForward 2, a joint program between DOE’s Office of Science and the National Nuclear Security Administration, is designed to develop affordable and energy efficient technologies to help in the pursuit of exascale computing over the next decade. The effort involves computing industry leaders like AMD, Cray, IBM, Intel, and NVIDIA.
A Bridge to Exascale?
Exascale-level supercomputers would be about 1,000 times faster than today’s machines, which are capable of speeds in the petaflop/s range, which is about a quadrillion operations per second. With an architecture that is viewed as one of the most promising pathways for exascale computing, NNSA officials say Sierra will serve as a key milestone in the pursuit of exascale computing when it is expected to come online in 2017. “Beginning in 2017, Sierra will be a key tool for the three NNSA laboratories in pursuing predictive applications necessary to sustain the nation’s nuclear deterrent into the indefinite future without underground testing,” Livermore Principal Associate Director for Weapons and Complex Integration Charlie Verdon said in a statement. “In particular, the machine will be dedicated to high resolution weapons science and uncertainty quantification for weapons assessment. This work promises to advance the state of the art in simulation science to the benefit of the larger research community.”
Oak Ridge National Laboratory said its supercomputer, Summit, will focus on simulations involving combustion science, climate change science, energy storage and nuclear power. “Summit builds on the hybrid multi-core architecture that the OLCF [Oak Ridge Leadership Computing Facility] successfully pioneered with Titan,” Buddy Bland, Summit’s director at the OLCF. “The large, powerful nodes allow applications to achieve very high performance without having to scale to hundreds of thousands of Message Passing Interface (MPI) tasks. The combination of very large memory per node and the powerful IBM POWER and NVIDIA processors provides an ideal platform for data analysis as well as computation.”
