Post-Workshop Feedback Reveals 50% Can Apply Fortran Knowledge Gained within One Month
The Research Computing Services department organized a Fortran programming workshop as part of its efforts to help researchers develop high-performance computing (HPC) skills and expertise that will enable them to accelerate their research.
The workshop was led by Dr. Wadud Miah, Scientific Computing Support, Khalifa University, HPC specialist and a computational scientist who uses modern Fortran to advance computational science. Starting with the structure of a Fortran code, Dr. Miah offered tips on ways to compile and link codes in the Linux operating system – the dominant operating system used in the Research Computing service.
Other topics included data types, printing and reading, control structures, procedures, modules, and pointers. The workshop also covered data management by presenting the NetCDF file format which allows users to store meta-data about their experiment and simulation, as well as share data with other scientists. Visualization which allows users to visualize their data and test their simulation was also covered. In addition, the workshop touched on parallel programming in OpenMP and Message Passing Interface (MPI) on how to parallelize codes, since larger and complex simulations require quicker time to solve.
The workshop had a strong emphasis on ‘applied’ to help participants apply their knowledge to accelerate their research. Practical exercises were given to reinforce participants’ knowledge and help clarify complex information.
Fortran remains the dominant programming language of HPC and accounts for around 70% of CPU cycles for the UK Archer supercomputer since many HPC applications in chemistry, physics, biology, environmental sciences, and materials science are developed in Fortran. The need for programming skills and expertise is essential for users who need to develop their own applications or extend the features of existing open-source applications, even though most Research Computing users use existing applications such as GROMACS, VASP, or Ansys.
Post-workshop feedback revealed that 50% will be able to apply the Fortran knowledge gained within one month and the remaining will be able to apply the knowledge between one and six months. Such promising statistics indicate how quickly the lessons learnt can be applied to research areas.
With an average rating of 4.75 out of 5, the workshop was also successful in terms of the extent of engagement by participants. Such dialogue leads to between the attendees which is highly encouraged as we would like attendees to collaborate and share their experience and knowledge with others.
The Research Computing Services places special emphasis on continuous learning and development to advance careers and accelerate science. Research Computing Services serves as a scientific instrument that helps scientists to be more productive.
If you have training ideas for a topic on Research Computing, please contact researchcomputing@ku.ac.ae
Further information on HPC workshops can be found here:
/research-
