Commit 76150277 authored by Alexandru Dura's avatar Alexandru Dura

Update the lecture notest

parent c1e5abb7
......@@ -52,3 +52,26 @@
* <2020-02-13 Thu>
- SIMD optimizations
* <2020-02-25 Tue> OpenMP tasks
* <2020-02-27 Thu>
** MPI
- Aurora nodes are managed by SLURM
- run only with exclusive access to nodes
- mpiexec --bind-to core ./processor_mpi
** Point-to-point communication
- eager vs rendezvous data transfer protocolos
- use synchronous send for development, then move to normal send to make code faster
Question: analysis tools for avoiding race conditions in OpenMP, or avoiding deadlocks in MPI
- there is no widely accepted answer;
- people are mostly relying on dynamic analysis tools, like Scalasca
- JH is available for more discussions in the autumn semester
* <2020-03-03 Tue>
- ARM DDT parallel debugger
- ~module load arm-forge~
- ~ddt~
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment