Manufacturing

In the manufacturing industry, HPC methods are used to design, optimize and test new products and refine existing products. Computing systems and the models that run on them help manufacturing researchers make their companies more competitive and profitable. Common computational methods include fluid dynamics, structural and materials science, basic and applied chemistry and process simulation.

Vehicle manufacturers use high performance computing to improve fuel efficiency, automotive design, and passenger safety. The multitude of variables involved in simulations such as automobile crash testing requires both a large number of processors and a high level of performance. Cray systems are built to run codes like the popular LS-DYNA software package, used for crash simulations and other complex, nonlinear physics-based problems.

Businesses in aerospace, electronics manufacturing, industrial design and materials production rely on computer-generated simulations to design and fine-tune their products.

The development of Boeing’s 787 Dreamliner involved 800,000 processor hours of computing time on Cray supercomputers. Virtual prototyping allowed Boeing engineers to build the aircraft after physically testing only 11 wing designs, versus 77 wing designs for the earlier Boeing 767, and resulted in a plane both lighter and with fewer emissions than similarly-sized craft.

Golf equipment manufacturer PING used a Cray system to model the performance of golf club shafts and optimize their design. The simulations analyzed the impact of a club head against a ball, bending of the shaft during a stroke, the stability of a putter while in motion, and other complex problems.

Virtual product development enables companies to save time, expense, and resources. Instead of using a traditional development process that involves early production and testing, creating virtual models eliminates these costly stages of physical fabrication and retooling.

The Cray XE series provides the unmatched scalability needed to push these models to their limits, increasing the accuracy and impact of the research on the product development cycle. The Cray XE6, using standard parallel programming paradigms such as OpenMP and MPI, offers a portable platform for both development and production simulations. 

The Cray CX1 deskside system, with powerful compute, IO and visualization nodes is specifically geared to the needs of manufacturing R&D, providing both a development and production simulation environment.

Startup Saves $7 Million, Cuts Time to Market by 83% with High Performance Computing