International HPC Summer School 2016
Parallel programming: classic track (MPI & OpenMP) 2
David Henty, PRACE, XSEDE, RIKEN, Compute Canada
June 2016
Slide contents
Message Passing Programming
Overview
Modes
MPI_Ssend
MPI_Bsend
Notes
MPI_Bsend
Notes
Send
MPI_Send
Solutions
MPI_Ssend
Solutions
Parallel Programming
Traffic Flow
Simple Traffic Model
traffic behaviour
Traffic simulation
State Table
Pseudo Code
how fast can we run the model?
a parallel traffic model
The Model
Pseudo Code
Message-Passing Strategy (1)
Parallelisation Strategy (2)
Parallelisation
Pseudo Code
a parallel traffic model
Non-Blocking Communications
Deadlock
Completion
Blocking Operations
Non-Blocking Operations
Non-Blocking Operations
Non-Blocking Communications
Non-Blocking Send
Handles used for Non-blocking Comms
Non-blocking Synchronous Send
Non-blocking Receive
Blocking and Non-Blocking
Communication Modes
Completion
Multiple Communications
Testing Multiple Non-Blocking Comms
Combined Send and Receive