Breaking records in computing benchmarks is nothing new to Red Hat - we’ve shown repeatedly that Red Hat Enterprise Linux can break the mold when it comes to the standards of computing. Today, we’re pleased to extend these accomplishments to the world of supercomputing, with Red Hat Enterprise Linux powering two of the top three supercomputers in the world, according to the Top500 list. Overall, Linux underpins the entirety of the 500 fastest supercomputers in the world.
Both machines were designed as a part of Department of Energy’s (DOE) CORAL initiative to be used for the most demanding scientific and national security simulation and modeling applications and enable U.S. leadership in computing.
Summit, announced in early June, is housed at the Department of Energy’s Oak Ridge National Laboratory (ORNL) and is, as of today, the fastest supercomputer in the world. Constructed of building block-type architecture, Summit is comprised of IBM POWER9 processors, NVIDIA Volta V100 GPUs, and Mellanox Infiniband, with Red Hat Enterprise Linux making the raw power of the world’s fastest supercomputer available to researchers and end users through a common operating system interface. Using a combination of artificial intelligence and machine learning capabilities, Summit is designed to power a variety of scientific applications, from cancer research to astrophysics. The fastest supercomputer in the world is also one of the most power efficient. According to Green500, which keeps track of energy usage by Top500 supercomputers, Summit ranks in the top ten globally.
Sierra is still being configured at Lawrence Livermore National Laboratory (LLNL) and already registers as the third fastest computer in the world. Scheduled to be fully accepted in August 2018, Sierra is designed to support the National Nuclear Security Administration as well as two additional laboratories - Los Alamos and Sandia - and brings to bear extensive capabilities in machine learning, high-resolution modeling, and simulation. Like Summit, the design of this supercomputer is based on similarly-configured building blocks with the world’s leading enterprise Linux platform acting as the common interface to compute nodes built from IBM, NVIDIA, and Mellanox technologies.
In collaboration with our partners, we’re helping to reshape the next generation of supercomputers to be more open and to support scientific advancement by providing a level of stability and support required for running mission-critical workloads. What’s more, these machines are all built with enterprise-grade hardware and software components at scale, providing a proof point for how the datacenter of the future can evolve to support advanced computing outside of scientific research.
Congratulations to ORNL and LLNL, to our partners, and to the Red Hat engineering and product teams who helped to drive these advancements forward!
About the authors
Yan Fisher is a Global evangelist at Red Hat where he extends his expertise in enterprise computing to emerging areas that Red Hat is exploring.
Fisher has a deep background in systems design and architecture. He has spent the past 20 years of his career working in the computer and telecommunication industries where he tackled as diverse areas as sales and operations to systems performance and benchmarking.