CFD lies at the core of most industrial design cycles in hydraulics, aerospace, casting, microfluidics and other engineering applications. As hardware architectures evolve, engineers are constantly trying to solve large complex simulations with reduced runtimes in order to optimize their design cycles. Flow Science developers have collaborated with Intel technical staff for over 7 years to improve the accuracy and performance of FLOW-3D using Intel’s constantly evolving architecture and toolsets to meet its customers demand for accurate solutions in less time for larger data sets.
"Working with the Intel® support team, we were able to exceed what had been a previous scaling limit of 32 cores. After enabling our software for distributed memory parallelism, we were able to run on more than 32 cores of Intel®’s latest processor architecture and find out how the software performed," said Anup Gokarn, Senior FLOW-3D/MP Developer.
Earlier this year, Flow Science collaborated with Silicon Mechanics, a leading manufacturer of rackmount servers, storage, and high-performance computing clusters, to offer a turnkey cluster solution for FLOW-3D/MP customers. The cluster solution features the most advanced technology from Intel: pre-installed FLOW-3D/MP; an exclusive on-line configurator;
The turnkey cluster solution developed for FLOW-3D/MP is an Intel® Cluster Ready-registered application and has been validated on Intel® Xeon® processor Product Family hardware, thereby ensuring that users enjoy the same solutions they’ve come to expect from workstations. Users are guaranteed that the hardware and software will work correctly together.
Photos:
http://www.prlog.org/
http://www.prlog.org/
