Episode 4.9 - Distributed-memory Parallelism and MPI
Table of Contents:
00:00 - Three layers of parallelism
00:28 - Distributed Memory Parallelism and MPI
00:40 - Explaining MPI
01:00 - Programmer defines role of MPI messages
01:31 - Portability of MPI applications
01:38 - Future-proofing of MPI aplications
01:49 - Compiling MPI application
02:09 - Running MPI application
02:32 - Structure of MPI application
02:41 - MPI_Init
02:52 - MPI_Comm_size
02:59 - MPI_Comm_rank
03:18 - MPI_Get_processor_name
03:23 - MPI_Finalize
03:29 - Passing Messages
03:37 - p2p messages (point-to-point communication)
03:45 - MPI_Send
03:49 - MPI_Recv
03:58 - Example of the message
04:03 - Function arguments
04:22 - MPI communicators
04:34 - Collective Communication
04:52 - MPI_Bcast
05:03 - MPI_Scatter
05:12 - MPI_Gather
05:22 - MPI_Reduce