PHYSICS

Intel Corporation and SGI today announced that the Procter & Gamble Company has recently doubled its deployment of Intel Itanium 2-processor-based SGI Altix systems used in the manufacturing and packaging design of some of its most popular consumer products. The products include Folgers coffee, Charmin bath tissue, Bounty paper towels, and Pampers diapers. The deployment helped save the beauty and health care product company time and money by enabling Procter & Gamble (P&G) to concurrently implement complex design projects. The Procter & Gamble Corporate Engineering Technology Group increased the Altix high-performance computing capacity to a total of 64 Itanium 2 processors and 96 GB of shared memory in January of this year. P&G's approximately 100-person design group is able to help eliminate costly and time-consuming physical product testing in developing its manufacturing equipment and product packaging by conducting more efficient product tests using "virtual" prototypes. "In just a few months we've seen dramatic increases in both the speed with which our engineers are able to work through the design process and the breadth of perspective they can apply to development projects," said Tom Lange, associate director of Corporate Engineering Technologies at Procter & Gamble. "For instance, by using the Altix system's global shared memory architecture to access product design data, we were able reduce the time of our stress analysis testing on the Folgers canister lid by a month. This wouldn't have been as possible nor as easy had we applied a traditional PC-based cluster solution to this problem." With help from the Itanium 2 processor-based SGI Altix system, P&G designers created the new Folgers AromaSeal* Canister. The structural challenges in replacing metal cans required significant computer-aided engineering (CAE) analysis and design input. In addition, CAE studies conducted on the Altix system also have helped designers achieve better-fitting Pampers diapers. The P&G team also uses the Altix system to run a series of "what-if" scenarios to help P&G determine how to get maximum manufacturing output for key products such as Charmin and Bounty. By testing a range of aerodynamic variables, engineers were able to understand the speed and stability limitations of processing paper at high speeds. Paper dust and moving sheets are complicated and expensive to manage and even more expensive to change based on guesses alone. The team used the results of computer modeling to evaluate settings and airflow management features in order to help maximize the processing rate while supporting quality and reliability. Prior to installing its initial 32-processor Itanium 2 processor-based Altix system, P&G only had enough computing power to run one complex job per person. With the newly expanded Altix deployment, P&G designers now have enough processing and memory capacity for each team member to run multiple what-if scenarios simultaneously. This enables them to make more informed design decisions related to product launches from the manufacturing process to final product packaging. The SGI Altix system architecture handles large data sets with ease, helping to enable customers to achieve groundbreaking improvements in manufacturing, oil and gas exploration, homeland security, earth and environmental sciences research, and life sciences. Since its introduction in January 2003, the Altix system has set numerous records for sheer performance, and for its ability to efficiently run manufacturing, engineering and scientific applications across hundreds of processors in a Linux* operating environment.