The aim of this course is to give users the best practices to go further using CURIE system and to give hints to prepare their codes for future architectures.
- architecture (hardware, interconnect, file-system),
- code optimisation (vectorisation, profiling),
- parallelisation (MPI, openMP, mixte),
- IO improvement.
code developpers, knowledge of C or F90, MPI, OpenMP
Day 1, tuesday 25
Introduction, Allinea tools
Current & future CPU architecture
Vectorisation & threads, OpenMP4
Day 2, wednesday 26
Intra nodes, profiling & tuning
Advanced MPI, point to point, MPI & threads, tuning librairies
Day 3, thursday 27
Advanced MPI, communicators, collectives, non blocking collectives,
Advanced IO, Lustre, meta data, Posix, MPI-IO, HDF5