WHO KNEW…..CUDA-ACCELERATED SHAMPOO!

by Hector Marinez

When deciding which shampoo to purchase, you usually don’t think of the research and development that went into its formula and design. The process for creating the perfect shampoo for consumers is more technically involved than one would think. Improving the cleaning power and organic compounds of liquid shampoos and detergents – while also ensuring they’re environmentally friendly – is as much a computer problem as it is a balance of chemicals.

To develop better surfactants – compounds that attach to dirt on hair and wash it away – a multinational household manufacturer needed a more powerful computing solution to better understand the real-time behavior of tens of millions of particles. The company enlisted the University of Pennsylvania, which partnered with NVIDIA, to help find a faster, more accurate and cost-effective solution to speed-up its simulation process. The goal was to develop more efficient wetting agents found in shampoo and household cleaners, as well as designing the bottles and packaging of those products.

“The massive simulations we are able to run on NVIDIA GPUs clearly show the superiority of the CUDA architecture and GPU Computing,” said Dr. Axel Kohlmeyer, associate director for the Center for Molecular Modeling. “We now have greater power and flexibility because of the GPU’s computing performance. We’re moving rapidly ahead to deploy a Tesla GPU cluster, which will give our work a huge boost.”

Dr. Kohlmeyer’s team immediately realized a tremendous increase in computing performance and eliminated exhaustive physical experiments and expensive lab tests, allowing products to reach store shelves much more quickly.  The computing capability of the GPU is no surprise to NVIDIA, but we’re proud of its ability to enhance the development of an ordinary household item such as shampoo.