[Sponsored Post] For data-driven researchers, the time-related expense of moving data from machines between data centers slows computation and causes costly delays in results. Plus, data center space can be limited for many organizations, even well-established academic campuses. To address these issues, Purdue University deployed NVIDIA InfiniBand MetroX across campus in 2011, connecting remote computation clusters to remote storage facilities.
Computing With MPI in Heterogeneous Environments
June 18, 2015 by
Designating the appropriate provider for large MPI applications is critical to taking advantage of all of the compute power available. “A modern HPC system with multiple host cpus and multiple coprocessors such as the Intel Xeon Phi coprocessor housed in numerous racks can be optimized for maximum application performance with intelligent thread placement.”
