Anycast: Rootless Broadcasting with MPI
Abstract: Flexible communication routines that enable high-performance computing (HPC) applications to operate with less synchronization are a great benefit to application algorithm creation and developer productivity. Further, communication operations that reduce synchronization in HPC applications while continuing to scale well are critical to application performance in the exascale era. We present an algorithm that allows an application to scalably broadcast a message from any MPI rank without specifying the origin of the message in advance, demonstrating application development flexibility, highly scalable operation, and reduced application synchronization.
Back to Workshop on Exascale MPI (ExaMPI) Archive Listing
Back to Full Workshop Archive Listing