Red Hat has partnered with several U.S. Department of Energy laboratories, targeting their efforts to progress cloud-native standards and practices in high-performance computing environments.
The Raleigh, North Carolina-based open source software company said Tuesday it will work with the research teams at Lawrence Berkeley National Laboratory, Lawrence Livermore National Laboratory and Sandia National Laboratories, among others, to expand the possibilities of harnessing standardized container platforms to connect HPC and cloud computing impacts.
Chris Wright, chief technology officer and senior vice president of Red Hat, outlined the prominence of HPC technology in demonstrating computer-intensive applications’ strengths in scientific experimentation via containers. Wright says this caused an imbalance of standards throughout HPC sites and a resultant difficulty in constructing and implementing new programs that can be used in HPC, commercial and cloud settings.
“Through our collaboration with leading laboratories, we are working to remove these barriers, opening the door to liberating next-generation HPC workloads,” Wright stated.
A widespread, growing desire to operate containerized syllabi on traditional HPC architectures is attributable mainly to new discoveries in artificial intelligence, machine learning and deep learning, in addition to compute and data-driven analytics.
In their joint ventures with the DOE, Red Hat looks to combine their expertise in cloud-native and multi-cloud infrastructure with the laboratories’ innate facility with large-scale HPC execution. The collaboration will center on four focus areas aimed at gaining new understanding in exascale computing, touching on standardization, scale, cloud-native application development and container storage.
The work will find Red Hat’s team utilizing some of the most intricate and robust supercomputers in the world. Their partnership with the National Energy Research Scientific Computing Center at Berkeley Lab will see the two organizations refine Podman, a daemonless container engine designed to develop, manage and run container images on a Linux system.
In addition, alongside the research staff at longtime partner Sandia National Laboratories, Red Hat will concentrate on deploying a Kubernetes-derived infrastructure at enlarged scale in order to test a simplified route for mechanism delivery of containerized workloads.
Red Hat’s new DOE announcement follows the company’s March deal to offer data integration services for a network of hospital research centers in Canada. Using Red Hat Integration, Toronto-based University Health Network consolidated 170 systems into one platform.