Thursday 5 March 2009

Day 3 in Catania

Yesterday I went to the computational chemistry session which consisted of 2 sessions each with 3 talks.

Compchem kicked off the proceedings – they are one of the EGEE VO’s (supported by several EGEE sites) and currently have over 60 users since they started at the end of 2004. Much like the NGS they also target user conferences in order to advertise their presence and the services they can offer the chemistry community. Again, like the NGS, they are perhaps restricted in which countries they can target and so attend European events. They also provide training for users to help them get up and running on the EGEE grid resources and to use popular software such as GAMESS, DL-Poly etc. It’s nice to see that in order to reach out to users that they have done case studies of previous research in order to highlight successes etc.
For me this presentation was the highlight of the first half of the computational chemistry session!

After the break I swapped sessions as it seemed that all the user orientated sessions ie those that actually featured real uses of grid technology, all clashed with each other.

The next session I went to was entitled “grid technology uptake by scientific communities and SME's”. As the user engagement person for the NGS I was obviously more interested in reaching out to the scientific communities rather than the SME's. There were a large number of rather short talks in the session although some of these had been cancelled.

Although Sicily might seem like a small island, it does have its own grid. The Sicilian grid consists of 200 cores, 2GB of f RAM per core, 250 TB storage and gLite 3.1 middleware. To manage the thorny issue of licensed software, they have a system to manager this as they have grid enabled their Flex server. It was fantastic to hear someone actually say that building infrastructure is a waste of time if you don't build a user base - hear, hear! I was very impressed to see that they have purpose built classrooms and training infrastructure in place to make sure that grid is seen as a common tool. Unsurprisingly they were able to show a direct correlation between training and usage of infrastructure.

No comments: