ACADEMIA

UK scientists building a computing Grid for particle physics have launched the next phase of their project, in advance of the start of the world’s largest experiment. Over the last six years, the GridPP collaboration has successfully built a distributed computer system for scientists working on the world's biggest experiment, the Large Hadron Collider (LHC) at CERN near Geneva. The Science and Technology Facilities Council (STFC), has extended the project for another three years which ensures that the expertise built up in the UK will be there for the start of the LHC later this year and for the crucial first years of data taking. The data crunching and storage capabilities of the Grid are essential to the LHC’s science mission of exploring the fundamental particles and forces of nature. The LHC is currently in the final stages of commissioning at CERN, the European particle physics laboratory in Geneva. Once switched on, expected to be in July 2008, it will collide sub-atomic particles inside cathedral-sized detectors, searching for clues to how the universe works. But to discover the next generation of physics, scientists will need to sift through the petabytes (100,000s of Gigabytes) of data that the experiment will produce. To store and analyse this data will require a whole new approach to computing - the world’s largest scientific computing Grid, a massive co-ordination of over 50,000 CPU in 50 countries, creating a virtual supercomputer that can be expanded and adapted. GridPP is the UK’s contribution to this Grid, with 10,000 CPU and 700TB (700,000 Gigabytes) of disk storage at 17 sites in the UK. The GridPP grid has already been helping scientists run simulations to prepare for the LHC, running over 7 million computer programs in 2007. Dave Britton, GridPP's new project leader is looking forward to the future of the project “I am excited to be leading the project as we move into the phase that will start to produce physics. The infrastructure that we have been building up will now be required to process huge amounts of information produced at CERN. With the LHC turn on just months away there is intense excitement, and a certain amount of understandable trepidation, amongst the collaboration but I am sure we will overcome the challenges that are going to be presented in the coming years. ” GridPP has been developing the system in conjunction with international partners and has been a leading force in the development of the Grid. In the UK the Grid has been a fully functional system for a number of years helping out other sciences and has recently passed the first computing readiness challenge for the LHC and is preparing for the second. The ultimate test for the group can only come from handling real data in real time. Tony Doyle has been Project Leader since the beginning of GridPP “It has been 7 years since we first bid to start work on the computing challenge that the LHC would present. We have come a long way since then and have achieved our goal of a large-scale production computing system for particle physicists in the UK. This next phase of the project will build upon this success as we enter the exploitation phase providing the necessary computing and data services required for LHC and other user analyses.” GridPP's main objective is to allow British scientists to process the data they need from the LHC to ensure that the UK is at the cutting edge of this important and ground-breaking new work. To this end, GridPP works very closely with these scientists to ensure their needs are being met. Roger Jones, Chair of the International Computing Board for ATLAS, one of the biggest experiments at the LHC, is glad to see that the UK has provided continued funding. “The ATLAS collaboration in the UK is delighted that GridPP has again been funded by STFC. Without this infrastructure, LHC physics in the UK would be almost impossible. I am very confident that the project is ready to deal with the impending challenges of the next few years and that physics, and science, in the UK will benefit from their experience.” Grid computing, which was originally conceived by scientists in the US, does not replace the Internet or the Web, but relies on both.