Generated by DeepSeek V3.2| Franklin (supercomputer) | |
|---|---|
| Name | Franklin |
| Location | Lawrence Berkeley National Laboratory |
| Manufacturer | Cray Inc. |
| Architecture | Cray XT4 |
| Operating system | UNIX |
| Memory | 20 terabytes |
| Speed | 852 teraflops |
| Ranking | 6th (TOP500, November 2008) |
| Purpose | Computational science |
| Decommissioned | 2012 |
Franklin (supercomputer). Franklin was a Cray XT4 supercomputer system operated by the National Energy Research Scientific Computing Center (NERSC) at the Lawrence Berkeley National Laboratory in California. Named for American scientist Benjamin Franklin, it was a major resource for United States Department of Energy (DOE) open science, enabling breakthroughs in fields like climate modeling, materials science, and astrophysics. From its deployment in 2006 until its decommissioning in 2012, Franklin served thousands of researchers and consistently ranked among the world's most powerful supercomputers.
Franklin was installed at the Lawrence Berkeley National Laboratory as a primary production system for the National Energy Research Scientific Computing Center (NERSC), a United States Department of Energy Office of Science user facility. Its primary mission was to support unclassified computational science research across a vast portfolio of DOE projects. The system was named to honor Benjamin Franklin, reflecting a tradition at NERSC of naming supercomputers after pioneering American scientists. Franklin succeeded the Cray XT3 system known as Bassi and was itself eventually succeeded by newer systems like Hopper and Edison.
Franklin was based on the Cray XT4 massively parallel architecture, a direct evolution of the earlier Cray XT3 platform. The system initially comprised 9,660 AMD Opteron compute nodes, each containing a dual-core central processing unit and connected via Cray SeaStar interconnect technology. A major hardware upgrade in 2008 expanded Franklin to 19,320 nodes utilizing quad-core AMD Opteron processors, significantly boosting its performance. This enhancement brought its total memory to 20 terabytes and its theoretical peak performance to 852 teraflops, a metric that earned it 6th place on the November 2008 TOP500 list of the world's most powerful supercomputers.
Franklin ran the Cray Linux Environment, a specialized UNIX-like operating system optimized for the Cray XT series. It supported a wide array of scientific software and programming models, including MPI and OpenMP for parallel computation. Key application areas that dominated its cycle usage included climate modeling projects like the Community Climate System Model, combustion simulations for engine design, computational chemistry codes such as NWChem, and astrophysics research studying supernovae and dark energy. The system also hosted large-scale materials science calculations and research in high-energy physics, including analyses for experiments at the Large Hadron Collider.
Franklin was accepted for production use at NERSC in 2006, following a procurement process managed by the United States Department of Energy. Its installation was part of a sustained effort by the DOE Office of Science to provide cutting-edge computational resources to the national research community. The major 2008 upgrade kept Franklin competitive with emerging supercomputing technology worldwide. After six years of highly reliable service, Franklin was officially decommissioned in 2012. Its retirement was part of the planned technology refresh cycle at NERSC, making way for its successor, the Cray XE6 system named Hopper.
Franklin had a profound impact on computational science, enabling peer-reviewed discoveries across numerous scientific disciplines. It provided essential computing power for projects that led to advancements in understanding climate change, designing new catalysts, and modeling complex physical systems. The system demonstrated the effectiveness of the Cray XT architecture for large-scale open science workloads and helped refine best practices for operating massive parallel computing environments. Franklin's operational success solidified NERSC's role as a leading facility for scientific computing and directly informed the design and procurement of subsequent supercomputers like Cori and Perlmutter at Lawrence Berkeley National Laboratory.
Category:Supercomputers Category:Cray computers Category:Lawrence Berkeley National Laboratory Category:History of computing in the United States