GridPP: Thoughts from Tampa at Supercomputing '06

GridPP made its annual trip to Supercomputing earlier this month. This year's conference was held in Tampa, Florida, where Dave Colling, Gidon Moont and Roger Jones presented the Real Time Monitor on the UK e-Science stand. Below, they tell us about why they went, whether they were asked any good questions and why GridPP had the best give-aways. Why did you (personally) go to Supercomputing? Gidon: To show the Real Time Monitor to people, and to see what other people were doing, particularly in Grids. Roger: To present what we are doing in the LHC Grid deployment and to talk with like minded people on other stands. Interestingly, when walking around I initially spent most of my time with the manufacturers, but this time I spent more time talking with people about applications. Dave: Good question. We have always produced the demo and so I went along to describe it to other people. What was GridPP doing there? Gidon: GridPP was showing how it contributes to the WLCG/EGEE projects, and highlighting these contributions as potentially more widely applicable in other projects. Roger: To show we are a key part of the UK e-Science activity, and to fly the flag for the UK. What were people interested in on the stand? Gidon: The RTM! And other stuff I am sure... Roger: I genuinely think we had more than out fair share of people coming to the RTM. The Mappa Mundi (Interoperation World Map) attracted less interest, although we did have someone come over after the press release. We certainly had more traffic than some others on the stand. This is partly because the monitor is dynamic and eye-catching, and partly because we dragged people in like circus hucksters. (Actually, that might be polite for some of the lines I used!) What were the most interesting questions you were asked, and the most interesting people who came by? Dave: The questions are the same every year. Is this really live? Can we use this software? Roger: We had some interest from companies like General Electric and HP, and other Grid deployments. We also had some computer scientists come over interested if there were job openings in LCG/EGEE. People were interested in the workload management and the matching of jobs to data (so am I; our answers were sometimes a little idealised). They also asked how we deal with different operating systems, and the usual, 'This is a simulation, right?' questions. What were the best stands you visited - and the best freebies you picked up? Gidon: No brilliant freebies - I suppose the BAE Systems organising folders on the stand next to us will be useful, but I think our laser pointers were cooler than most other stuff. Roger: Getting free drinks off Microsoft at the Microsoft party is always something for your diary. I was interested in some finite volume weather simulations on (I think) the Intel stand, partly because it is related to some non-physics projects I have an interest in. (Coastal dynamics, since you ask). Roger and Dave gave talks on the UK e-science stand, entitled, "A Massive Undertaking: From Grids to the Higgs Boson, applications for particle physics on the LCG/EGEE Grid", and "GridPP: Running and monitoring a production Grid". These were complemented by talks from Simone Campana and James Casey of CERN, who talked about application-middleware integration, and building the LCG. Other GridPP members on the UK stand included John Gordon and Pete Clarke. Pete can be heard on CERN's Gridcast web pages talking about e-Science participation in SC'06 its Web site CERN also ran a blog during the conference, which gives a daily insight into life on their stand: its Web site. Demonstrations on the UK stand are chosen each year by the UK e-science programme. If you are working on a demo that GridPP could present next year, please contact Sarah Pearce (s.pearce@qmul.ac.uk). Credit: GridPP