One important goal of PIC is the exploitation of GRID methodologies and technologies to enable optimal and extensive use of global scale distributed computation, to endow scientific communities with the processing and management data resources they need for an elite research.
Advances in microelectronics, scientific instrumentation and methodology, together with the exponential grow of the capabilities of computers, have strengthened the digitalization of all kinds of experimental results and theoretical simulations. In fact, leading science has gone even further and, nowadays, there are research centres that generate an amount of data that exceeds their local resources. Furthermore, often sharing data between various groups at a global level can benefit the objectives, common or singular, of all them. The GRID provides a solution on both ways.
The potential of centres with GRID resources, such as PIC, lies in the support for collaborations between different institutions for shared use of scientific data in a secure manner, in a way that optimizes the performance of the computing resources of the institutions involved. Adapting the advances in the research and computer technology domains about GRID, the PIC allows processing at an affordable price of large amounts of data, in order to promote the immediate use in other subjects that can benefit from them.
Maybe this scientific approach may seem to position PIC and GRID technologies far from everyday needs. Some further thought can lead to a different conclusion, however. Remind yourself of the beginnings of the latest large innovation in the computing world, Internet: scientists were the first to intensively use it, but nowadays nobody can deny that this technology has arrived to all corners of our society and that, in fact, has had considerable impact on it.
The GRID philosophy can be applied at small and large scale, from projects involving a small number of centres up to projects on European scale and beyond. The scope of PIC's activities has been set to address support for scientific communities that need technological innovation in order to benefit from GRID technologies in subjects which need treatment of massive amounts of data under extremely difficult conditions.
Some examples of such communities are found in the application examples within GRID technology development projects with National or European financing, such as the DataGrid, the CrossGrid and Enabling Grids por E-sciencE projects in which IFAE and UAB are participating:
- High Energy Physics, which is building at CERN, the European Organization for Nuclear Research, the Large Hadron Collider (LHC). When this accelerator starts its operation, between 5 and 8 million GigaBytes per year have to be stored and analyzed from different points of the European, American and Asian geography.
- The studies of the human genome, which is only one of the many genomes that have been sequenced. An instrument to understand how the human DNA works is to compare it with that of other species, which implies traversing many times over thousands of GigaBytes information, searching for non-predictable patterns.
- The European Space Agency (ESA), with its project of Earth intensive monitoring. ENVISAT generates about 500 GigaBytes per day in Earth images for complex studies such as climate change.
- The MAGIC collaboration operates a two 17-m telescope system for gamma-ray astrophysics. The amount of data generated by MAGIC amounts to 100 TeraBytes per year and will increase in the following years to 400 TeraBytes per year. This data has to be stored, analyzed and made available to all collaboration members. Also the generation of a Monte Carlo library of simulated gamma-ray events is an important task for the collaboration which demands lots of cpu power.
- New projects in astrophysics such as PAU (Physics of the Accelerating Universe) or DES (Dark Energy Survey) use simulations extensively to test their futue experiment results. These simulations, generated in supercomputers, need to be available to scientists.
In all these cases, we are dealing with research that needs to analyze a large amount of data and that, moreover, have to be shared between scientists in different locations.
Apart from these communities, which have already adopted GRID computation for their studies, we can develop others which are just around the corner. An example is the collaboration among different hospitals to share, in a secure and anonymous way, their data -to go forward in the prevention, detection and treatment of breast cancer for example. If you give freedom to your imagination, there are many other examples of emerging data-intensive collaborative science.
