"We are thrilled that CD-adapco has become part of the growing ecosystem of independent software vendors that support the PathScale InfiniPath InfiniBand technology," said Scott Metcalf, CEO and president of PathScale. "By driving messaging rates well beyond that of any other interconnect technology, InfiniPath will enable STAR-CD users to scale their simulations to new heights and decrease their time-to-solution for critical research, ranging from automotive and aerospace design to environmental science."
The InfiniBand standard is quickly becoming the networking architecture of choice for organizations that are looking to deploy large clusters of commodity servers. Native support for PathScale InfiniPath InfiniBand interconnects lets STAR-CD users employ commodity AMD or Intel-based servers, instead of expensive, proprietary Unix-based systems, to run highly complex CFD simulations. This provides substantial cost of ownership benefits to a wide range of engineering fields, including aerospace, automotive, biomedical, construction, chemical, environmental, marine, power generation and turbo machinery.
"With the InfiniBand interconnect becoming more prominent every day, it's critical that our solutions support the technologies that help drive application scaling and overall performance on Linux clusters," said Steve MacDonald, president and founder of CD-adapco. "The performance demonstrated by PathScale's InfiniPath InfiniBand interconnect will provide STAR-CD users a new approach to scale their most complex CFD simulations -- at a very attractive price point."
The key to InfiniPath's tremendous performance advantage is the InfiniPath HyperMessaging Architecture, which delivers the highest message rate and highest effective bandwidth of any cluster interconnect available. It allows InfiniPath to support more than 10 million messages per second, or 10X-MR, more than 10 times the message transmission rate of any other cluster interconnect. As a result, STAR-CD users will significantly increase the scalability of their applications, reduce network fabric congestion and improve overall cluster efficiency.
|