Parallel Linear Algebra (PRACE training course)

2 Feb 2017, 08:30 → 3 Feb 2017, 16:00 Europe/Prague

auditoriums NA4, NA5 (VŠB - Technical University Ostrava, building NA (New Aula))

Annotation

The course Parallel Linear Algebra (PLA) focuses on modern parallel methods for solving large linear systems of equations on parallel computers.

It is organized as a joined event of the National Supercomputing Center IT4Innovations and the French Maison de la Simulation, which is one of the six PRACE Advanced Training Centres, providing high-level training for the Partnership for Advanced Computing in Europe (PRACE). Moreover, it will be co-located with the traditional Seminar on Numerical Analysis and Winter School 2017 (SNA), held on 30 January - 3 February 2017.

Detailed information, including agenda and registration, is available on the PLA web page on the PRACE events portal

Level

intermediate-advanced

Language

English

Acknowledgements

The course is supported by PRACE and by the Ministry of Education, Youth and Sports from the programme Large Infrastructures for Research, Experimental Development and Innovations as part of the project „IT4Innovations National Supercomputing Center – LM2015070“.

Course materials

• PLA-02-1017-MdS_PATC.pdf
• PLA-02-2017-M.Faverage-1-DenseLA.pdf
• PLA-02-2017-M.Faverage-2a-SparseDirect.pdf
• PLA-02-2017-M.Faverage-2b-SuperNodal.pdf
• PLA-02-2017-M.Faverage-3-HybridSolver.pdf
• PLA-02-2017-Z.Strakos.pdf

Links

• 0 Original web page of the course
• 1 PLA web page on the PRACE events portal
• 2 SNA web page (in Czech)
• 3 Maison de la Simulation
• 4 Partnership of Advanced Computing in Europe
• 5 Anselm and Salomon documentation
• 6 Accommodation and transport recommendations (in English)
• 7 Transportation and accommodation - info for distant participants (in Czech)

Thursday, 2 February

08:30 → 09:00 Registration

09:00 → 09:45 IT4Innovations - intro and IO hardware overview (Branislav Jansík, IT4Innovations)

09:45 → 10:45 General parallel IO strategies (Sebastian Lührs)

10:45 → 12:00 lunch break

12:00 → 12:45 IO Profiling (Sebastian Lührs)

12:45 → 13:30 MPI-IO (Nicole Audiffren); Overview of MPI-IO (motivation, file view code, collective I/O, hints)

13:30 → 14:00 coffee break

14:00 → 15:30 MPI-IO (Nicole Audiffren); Performance and ROMIO (hints)

15:30 → 16:00 coffee break

16:00 → 17:00 MPI-IO (Nicole Audiffren); Hands-on

16:00 Coffee break

Friday, 3 February

08:00 → 09:30 HDF5 (Nicole Audiffren); Overview of basic HDF5 and API (manipulating HDF5 files, predefined datatypes, dataspaces and datasets, writing & reading data); Hands-on (serial HDF5)

09:30 → 10:00 coffee break

10:00 → 11:30 Parallel HDF5 (Nicole Audiffren); Overview (creating/accessing a file, writing and reading hyperslabs programming model); Hands-on: hyperslab example

11:30 → 12:30 lunch break

12:30 → 14:00 SIONlib (Sebastian Lührs); Introduction (motivation, SIONlib file format); Basic routines; Hands-on

14:00 → 14:30 coffee break

14:30 → 16:00 SIONlib (Sebastian Lührs); Advanced SIONlib features; Utility tools; Hands-on