The CHEP series of conferences, held every 18 months, covers the wide field of computing in high-energy and nuclear physics. CHEP'09 was the 17th meeting and attracted 615 attendees from 41 countries. It was held on 23–27 March in Prague. The conference was co-organized by CESNET, Charles University in Prague – Faculty of Mathematics and Physics, Czech Technical University, the Institute of Physics and the Nuclear Physics Institute. Throughout the week some 560 papers and posters were presented. As usual, given the CHEP tradition of devoting the morning sessions to plenary talks and limiting the number of afternoon parallel sessions to around six or seven, the organizers found themselves short of capacity for oral presentations. This time 500 offers were received for 200 programme slots with the rest being shown as posters, split into three full-day sessions of around 100 each day. The morning coffee break was lengthened to permit the attendees to browse the posters and discuss them with the authors.
Given the timing of the event, a large number of the presentations related to computing for the LHC experiments but there was also a healthy number of contributions from experiments taking place elsewhere in the world, including the US labs BNL, Fermilab and SLAC (where BaBar is still analysing its data although the experiment has stopped data-taking), KEK in Japan and DESY in Germany.
The conference was preceded by a Worldwide LHC Computing Grid (WLCG) workshop, which was summarized by Harry Renshall (CERN). There was a good mix of Tier 0, Tier 1 and Tier 2 representatives. It started with a review of each experiment's plans, all of which include more stress testing in some form before the restart of the LHC. EGEE to EGI transition is an issue, as is the lack of a winter shutdown in the LHC plans. The workshop summary was that ongoing emphasis should be put on stability, preparing for a 44-week run and continuing the good work on data analysis.
Sergio Bertolucci, CERN director for research and scientific computing, gave the opening talk of the conference, reviewing the LHC start-up and initial running, the steps being taken for the repairs after the incident of 19 September and how to avoid any repetition, and the plans for the restart. He compared the work being done currently at Fermilab and how CERN will learn from this in the Higgs search. Les Robertson (CERN), who led the WLCG project through the first six years of its life, discussed how we got here and what's next. A very simple grid was first presented at CHEP in Padova in 2000 and he labelled the 2000s as the decade of the Grid. Thanks to the development and adoption of standards, grids have developed and matured, and an increasing number of sciences and industrial applications have made use of them. But Robertson thinks that we should now be looking at locating Grid centres where energy is cheap, using virtualization to share processing power better and starting to look at clouds.
The theme of using clouds came up several times later in the meeting, for example the Belle experiment at KEK is experimenting with the use of clouds for Monte Carlo simulations in its planning for SuperBelle; and the STAR experiment at BNL (Brookhaven) is also considering using clouds for Monte Carlo production. Another of Robertson's suggestions for future work – virtualization – was one of the most common topics throughout the week in terms of contributions. Different uses of it cropped up time and again in multiple streams.
Among the other notable plenary talks was that by Neil Geddes (STFC, Rutherford Laboratory) who asked "can WLCG deliver?" and deduced that it can, and it does, but that there are many challenges left to face. Kors Bos (ATLAS) compared the different approaches to computing across the LHC experiments, pointing out similarities and differences. Ruth Pordes (Fermilab), executive director of the Open Science Grid (OSG), described work happening in the US with regard to evolving grids, making them easier to use and more accessible to a wider audience.
The conference had a number of commercial sponsors, in particular IBM, Intel and Sun Microsystems, and part of the Wednesday morning was devoted to speakers from these firms. IBM used its slot to describe a machine that it says offers cooler, denser and more-efficient computing power. Intel focused on an effort to get more computing for less energy, noting work done under the openlab partnership with CERN. Intel hopes to partially address this by increasing computing energy efficiency (denser packaging, more cores, more parallelism) because they realize that power is constraining growth in every part of computing. The Sun speaker presented some ideas on building state-of-the-art data centres. He claims that raised floors are dead – he proposed "containers" or a similar pod architecture that has built-in cooling and a modular structure connecting to overhead hot-pluggable busways. Another issue is to build "green" centres and he quoted solar farms in Abu Dhabi and a scheme to use free ocean cooling for floating ship-based computing centres.
It is impossible to summarize the seven streams of material presented in the afternoon sessions but some highlights deserve to be mentioned. The CERN-developed Indico conference tool was presented and statistics showed that it has been adopted by more than 40 institutes and manages material for an impressive 80,000 events. The summary of the 44 Grid middleware talks and 76 poster presentations was that production grids are here, Grid middleware is usable, standards are evolving but have a long way to go, and network bandwidth use seems to keep pace with technology. From the Distributed Processing and Analysis stream of talks came the message that a lot of work has been done on user-analysis tools since the last CHEP, with some commonalities between the LHC experiments. Data management and access protocols for analysis are a major concern and the storage fabric is expected to be stressed when the LHC starts running.
Dario Barberis (ATLAS) presented the conference summary. He had searched for the most common words in the 500 submitted abstracts and the winner was "data", sometimes linked with "access", "management" or "analysis". He noted that users want simple access to data so we need to provide easy-to-use tools to hide the complexity of the Grid. Of course "grid" was another of the most common words but the word "cloud" did not appear in the top-100 although it was much discussed in plenary and parallel talks. For Barberis, a major theme was performance, at all levels from individual software codes to global Grid performance. He felt that networking is a neglected but important topic (for example, the famous digital divide and end-to-end access times). His conclusion was that performance will be a major area of work in the future and a topic at the next CHEP in Taipei on 17–22 October 2010.
Useful links
CHEP'09 trip report: http://cdsweb.cern.ch/record/1173073?ln=en CHEP'09 programme including presentations: http://indico.cern.ch/conferenceTimeTable.py?confId=35523
A version of this article was published in CERN Courier, July 2009.