Monday, September 18, 2017

Support of Intel Omni-Path HPC Interconnect

The Transient Solver is one of the flagship solvers of CST STUDIO SUITE®, and with MPI computing it can be used to simulate electrically very large and complex models efficiently – for example: lightning strike, large antenna arrays, or antenna placement. Typically, the solver loop/solver run is where most of the computational time is spent during a transient simulation. MPI computing allows the computational workload of a large model to be split across computer cluster with a high-speed interconnect. CST STUDIO SUITE 2017 adds support for Intel’s new Omni-Path HPC interconnect. The picture shows the strong scaling of the transient solver loop on cluster system consisting of dual Intel® Xeon E5-2697A v4 processors, 64GB RAM per node, and an Intel® Omni-Path interconnect. The model size is 200 million mesh cells.

No comments:

Post a Comment