announced today that
Emirates Team New Zealand
, two-time America’s Cup champion, utilized high-performance (HPC) computing technology to design the yacht that challenged Oracle Team USA in the 34
Faced with the challenge of a new multihull boat requirement, Emirates Team New Zealand turned to Dell to revolutionize their design process with a Dell
, which provided the power of computational modeling and the ability to glean the insights needed for success on the water.
Dell worked extensively with Emirates Team New Zealand to develop an HPC environment configured to their computing needs, optimizing performance without any wasted power. With a high-performance computing environment, the team was able to design and rapidly and repeatedly test new vessel design concepts without requiring actual build-out. Previous yacht development involved physical tank testing of scale models of individual hulls.
“We can now complete an entire boat design test in three days using the Dell HPC cluster. We’ve gone from 30 to 40 design candidates being tested physically for our 2007 Cup campaign to testing 300-400 designs for this edition of the America’s Cup,” said Nick Holroyd, Technical Director, Emirates Team New Zealand.
Smart boat design has always been a critical factor in determining the America’s Cup outcome, and this year’s competition raised the stakes even higher, instituting the new multihull model for competing vessels. The Kiwis decided to dramatically increase their compute power to handle the complex design and testing required for success. And Dell put that same focus on the design of the HPC that would design the team’s vessel, fine tuning the cluster for these very specialized workloads with the help of the
Dell/Cambridge HPC Solution Center
at Cambridge University in London.
“The transition to AC72 multihulls has meant a complete shift in focus to technology where all our prototyping is computer generated,” said Holroyd. “With less than a year to design, test and build a new design, performance and the design had to be derived computationally, which meant our requirements for a high-performance cluster suddenly exploded.”