I was thrilled when I recently heard that the Department of Energy (DOE) launched its new High Performance Computing for Manufacturing Program (HPC4Mfg). Led by Lawrence Livermore National Laboratory (LLNL), the program aims to advance high performance computing (HPC)—a technology critical to a more cost-effective and efficient manufacturing sector.
HPC—which includes capabilities like Computer Aided Engineering and complex simulation—combines multiple systems and taps into vast data networks to perform complex calculations in ways that were only previously achievable with supercomputers. In manufacturing, HPC is critical to enabling the optimization of materials, processes, and end products in the most efficient ways possible at a commercial scale.
HPC relies on a combination of materials science and computer science to make design, simulation, and manufacturing faster and more cost-effective. Materials science links materials structure, property, and performance information, while computer science uncovers complex linkages between materials to gain insight into better ways to design and manufacture materials and other end products.
While a number of initiatives have or are currently focused on advancing HPC (see information below), HPC4Mfg is unique in its intention to impact a broad range of industries. Because HPC4Mfg is housed at a national laboratory, the HPC technology it develops will be widely accessible to the entire manufacturing industry. Rather than targeting small-scale productions of specific materials, the focus of HPC4Mfg will be driven by proposals from industrial partners—particularly from energy-intensive sectors such as chemicals, food processing, glass and cement, petroleum refining, primary metals, and wood pulp and paper—to help a wide range of industries address significant manufacturing challenges at a commercial scale.
By increasing industry-wide access to the tools and equipment needed for HPC, there are a number of opportunities that HPC4Mfg has the potential to help accelerate, including the following:
- Design of more energy-efficient and higher-performing materials: According to the MGI Strategic Plan, the manufacturing sector must advance materials for energy storage systems, lightweight materials, and structural materials to achieve national objectives in national security, human health and welfare, clean energy systems, and infrastructure consumer goods (see Nexight CTO Warren Hunt’s blog post for more detail). Modeling and simulation using HPC can accelerate materials discovery processes and the market readiness of these materials.
- Advancement of big data and emerging R&D disciplines: One study shows that the amount of operational manufacturing data has increased 1,000 fold in the last 10 years. While many people would consider the use of data analytics in computer science to be old hat, it is actually a relatively new discipline and one that complements HPC. For this reason, high performance data analytics (HPDA) is likely to evolve in parallel with HPC in the coming years. According to an insideHPC blog post, the HPDA market is forecasted to grow at a rate of more than 20% CAGR (compound annual growth rate) from 2013 to 2018 while significantly intensifying new R&D disciplines, such as additive manufacturing.
As the development and use of advanced HPC increases through programs like HPC4Mfg, it will continue to be a powerful technology that spurs growth in basic science as well as applied engineering. While it’s hard to predict the next step in the continued evolution of HPC, it is my hope that the research and industrial communities can keep working together to develop innovative ways to maximize its benefits and revolutionize manufacturing as we know it.