|Topic:||When everything connects – the synergy of big science and ICT|
|Author:||Professor Rolf-Dieter Heuer|
Professor Rolf-Dieter Heuer is the Director General of CERN. His five-year mandate at CERN, which began in January 2009, covers the start-up phase of the laboratory’s new flagship facility, the Large Hadron Collider (LHC), whose physics research programme got underway in March 2010. Prior to CERN Professor Rolf-Dieter Heuer was the Research Director at DESY, Germany’s largest particle physics research centre. Professor Rolf-Dieter Heuer began his career working at the University of Heidelberg, working on the JADE experiment at the DESY laboratory in Hamburg. He then moved to CERN, where he held a number of key positions, culminating in becoming spokesperson for the OPAL collaboration, one of the laboratory’s flagship experiments. He later returned to Hamburg to take up a Professorship at the University until named to the Research Director post at DESY. Professor Rolf-Dieter Heuer earned a PhD in physics from the University of Heidelberg.
CERN has long been at the vanguard of ICT. CERN’s Tim Berners-Lee invented the World Wide Web to foster scientific collaboration; the Web brought the Internet to the public at large and changed our world. To process the colossal amounts of data generated by the Large Hadron Collider, CERN pioneered ‘grid computing’ – sharing massive amounts of processing power and data storage at institutes around the world. The Web changed information access and the world. Will the Grid do the same with computing?
When I first came to CERN in the 1980s, the Laboratory was a regional centre with a 30-year tradition of hosting large international collaborations. It still is today, but the definition of large has evolved. In the 80s, large meant maybe 100 people. Today, it means thousands. It’s perhaps this demographic shift that has done more than anything else to propel the science of particle physics into the digital age. CERN has always been in the vanguard of ICTs. In the 70s and 80s, our computer scientists strove valiantly to equip the lab with networks to replace the manual transport of 6250bpi tapes, and the control systems of our accelerators were computerized. These developments brought a number of developments that were well ahead of their time, like touch screens and trackballs, and it set the scene for Tim Berners-Lee to invent the World Wide Web at CERN at the end of the 80s, when collaboration sizes had grown to several hundreds and a new way of sharing information was needed. The Web brought the Internet to the public at large, and made it a platform for information exchange. Today, CERN again finds itself at the cutting edge of Internet technology, this time in developing Grid computing. What the Web did for information, the Grid does for computing resources – sharing processing power and data storage. It’s a nice little historical twist that this is precisely what the first computer networks were built for. In the 1960s when computers were large, specialized devices networks were designed to allow people to access them remotely. It was only later with the advent of personal computing, email and, ultimately, the Web, that the Internet became more dominantly used for sharing information. At CERN, the concept of Grid computing was an obvious choice for data handling and analysis for our current generation of experiments. Today’s collaborations involve hundreds of universities and thousands of individual members around the globe. The particle detectors they have built are of enormous complexity, with up to a hundred million readout channels. The purpose of these detectors is to study high-energy particle collisions that allow researchers to address fundamental questions about our Universe. Some of the phenomena the researchers will address are extremely rare, demanding an extremely high collision rate to maximize the detectors’ chances of finding them. All this has led to the construction of a particle accelerator of unprecedented complexity: the Large Hadron Collider, LHC. The LHC collides particles with about seven times more energy than its predecessors, and will reach around a hundred times more intensity in its beams, generating up to 600 million proton-proton collisions per second. This adds up to a colossal amount of data, and even after filtering the 600-megahertz initial collision rate down to around 200 hertz of data to store, it still equates to some 15 petabytes of data per detector per year. The solution CERN and its partners have put in place is called the Worldwide LHC Computing Grid, WLCG, and it federates computing power around several hundred institutes all over the world. CERN is the centre of the WLCG, with about ten per cent of the total CPU capacity and the ability to store all the data. CERN’s computer centre is connected to a number of large ‘tier 1’ computer centres via dedicated 10 gigabit per second optical fibre links, and these in turn connect across the Internet to well over a hundred ‘tier 2’ centres. To the user, however, it appears as if all this power is on the desktop. With the LHC running smoothly since March this year, the WLCG is proving itself well up to the task of serving the particle physics community, and it is also playing an increasingly important role in other areas of science. Particle physics is not the only field of research that can benefit from the Grid’s distributed computing infrastructure. As a consequence, particle physics is not only at the forefront of this exciting new field, but also finds itself in the spotlight of other sciences, and we’re actively working to share our know-how with them. With support from the European Union, CERN led the Enabling Grids for E-science project (EGEE) from 2001 to 2010. EGEE’s goal was to build on the Grid infrastructure put in place for the LHC to bring in other sciences, and it has been a resounding success with fields as diverse as drug discovery, detection of breast cancer through mammography, Earth observation and mineral exploration taking part. EGEE’s success is underlined by the fact that this activity has been spun-off to a new organization called the European Grid Initiative, www.egi.eu, which will coordinate a federation of national grid infrastructures in Europe for multiple sciences. To give just one example of an EGEE success story, the Wisdom project has used Grid computing infrastructure to identify molecules that could form the basis of new anti-malarial drugs, cutting down a process that normally takes 15 years to just three. After screening a million potential molecules, 30 were identified for further investigation and are now undergoing testing in the lab. Shared infrastructure naturally nurtures the kind of interdisciplinarity that EGEE has come to exemplify, and that’s something that particle physics is no stranger to. When CERN built a unique low-energy antiproton facility for fundamental physics studies in the 1990s, little did we know that a decade later it would be put to use to explore a novel approach to cancer therapy. And when the laboratory’s first large accelerator came on stream in 1959, no one suspected that it would eventually host a collaboration of particle physicists, climate and aerosol scientists making important measurements about the nature and formation of clouds. These are the kinds of serendipitous results that come from openness, shared infrastructures and patience. In the future, similar multi-disciplinary teams and shared computing infrastructure will be necessary to address a spectrum of major societal problems ranging from the supply of essential utilities such as energy and water to population demographics. Will grid computing find its way into our homes? It’s probably too early to say, but it certainly has the potential. If Grids can bring scientific data to scientists around the world, why not to school students or even interested lay people? CERN has already initiated a similar initiative called LHC at home, which allows people to use their PCs to carry out calculations of importance to the LHC. This approach to citizen-based science shows very clearly that there’s a very high level of public interest in getting involved and has led to the creation of a new Citizen Cyberscience Centre, hosted in the United Nations Institute for Training and Research (UNITAR) offices at CERN. Another way that Grid technology is already being felt in business is through cloud computing which builds on many of the technologies and approaches developed for grids. In science, we are so used to connections that we use them instinctively. All regions participate; all fields are welcome. The combination of science and ICT naturally promotes serendipity, interdisciplinarity and openness. Taken together, these ingredients add up to progress.