Chapter member Narine Gevorgyan participated in the  Advanced School on High Performance and Grid Computing held at the Abdus Salam International Centre for Theoretical Physics, Trieste, Italy, 11-22 April, 2011.

HPC and Grid Computing is having a major impact in many areas of physics, chemistry, biology, engineering; as well as in material, environmental, and social sciences. The traditional education for scientists who design, implement and use their scientific software is focused mainly on explaining the physical principles their codes are founded upon. For this reason, there is an ever-growing need to provide an equivalent background in computer skills, to maximize the scientists’ ability to use HPC/GRID technologies, and to improve their applications performance. This becomes more and more important with the emergence of novel computing architectures based on many/multi-core CPUs, the increased role of GPGPUs etc.

The goal of this advanced school is to provide young scientists and engineers already active in computational sciences with key additional skills necessary to identify and employ the right computing infrastructure to run their numerical simulations and solve computational intensive problems. Topics will include the GPGPU role in HPC, interoperability issues among different computational infrastructures, tools and libraries for scientific data management, design and installation procedures of medium size computational facilities.
The school will be held for a period of two weeks. The first week shall be dedicated to an advanced introduction to e-Infrastructure for HCP and GRID computing, with the emphasis on their practical use. Theoretical lectures will be combined with the practical exercises in a computer laboratory where students will practice the concepts discussed during the lectures. The second week is dedicated totally to the running of personal/and or group projects in our laboratory. Also a number of short seminar lectures on advanced topics shall be presented by HPC and GRID experts.
Students will work actively on their own specific computational problems, adapting them to the e-Infrastructure environment of their choice.