Interview with Steven Newhouse

At the end of last month's EGI Technical Forum 2012, we caught up with European Grid Infrastructure (EGI) director Steven Newhouse to get his take on the week's highlights, as well as the challenges for the year ahead.

"The challenge with infrastructure is always making it visible,"
says Newhouse.

What have been the highlights of the technical forum for you?

It’s been great to see some of the activities we started off over a year ago coming to fruition. For example, the Federated Cloud Task Force has brought together what is going to be the next generation of our infrastructure; it’s leveraging skills and expertise from the NGIs [National Grid Infrastructures] and directing it into supporting user-driven scenarios. It’s been brilliant. It’s taken a lot of investment over the last year to bring this together, so that’s really nice to see. Of course, there are other activities, but this for me is important as it allows us to support new user communities and their technologies.

Anything else...?

Another new activity we’ve launched this week is called ‘EGI champions’. We’ve really started to focus on activities beyond the production infrastructure and what we do to reach out to new communities, how we connect to and grow our current communities and undertake general outreach. The EGI champions scheme is all about building connections with research institutions and even individual research groups. By identifying EGI champions we can invest in them. We can support them with information about what we’re doing and the work that they do in talking to people within their own research field. Also, we can listen to them about what they see the challenges are from their experiences working with the grid on a day-to-day basis. This will not only allow us to hear about what we can do better,  it will also allow us to find out what the opportunities are in terms of deploying new technologies and supporting new researchers across the whole of Europe in what they’re doing. We aim to prototype the project over the next year, so that when future funding streams come online we can aim to have champions in every research group across Europe that are active in computing activities and data analysis. That will give us links across the whole of the European Research Area and will allow Europe to know about computing and e-infrastructure and the benefits that come with it for their research activities.

What about the negatives as well? Are there any challenges which have come to light this week which you weren’t previously aware of — or perhaps as aware of as you are now?

I think the biggest concern is, bluntly, the level of funding. For Europe to have a strong grid infrastructure, investment in national infrastructure is required. With the challenging economic conditions across Europe, the signals coming from some countries regarding future funding aren’t as positive as we would like. What’s obviously very challenging about EGI is that it has over 35 countries in it and different countries have different appreciations of what the economic impact is. For Europe and the European Commission not to have an e-infrastructure divide, it is essential that funding comes along to make sure that the countries that are less strong are able to strengthen their activities and be reinforced in this area. This long-term funding challenge is my greatest concern.

Newhouse says that one of the highlights of this year's technical
forum was the launch of the EGI champions scheme.

Is this concern something that has underpinned the development of the new EGI champions scheme?

Not specifically during the prototyping phase, no, but further investment will be needed to roll this out on a  full-scale across Europe. Everything we’re doing and have been doing for many years is about positioning us to be more relevant to a broader set of users and researchers across Europe. The champions scheme and other initiatives are just ways of adapting what we’ve done in the past to the changing technological needs and support needs of the community which exists now. Part of this is about us understanding those communities better, but it’s also about communities themselves changing in response to the different technologies available. It’s a continuously moving goal, in terms of what you need to do.

There has, of course, been much talk at the forum about the role the grid infrastructure played in the recent discovery of the Higgs boson. Do you think that the public, however, is generally aware of this?

The challenge with infrastructure is always making it visible. I would love the community and the work that goes on here to gain more recognition than it has done. But the grid, if not EGI by name, was mentioned by CERN’s director general. Rolf Heuer, as one of the top reasons why the Higgs discovery was made, so there’s been a clear acknowledgment of the function that we have and it’s clearly excellent news for us that the community’s work has been recognized.

Is there perhaps an issue though that the grid has now become too focused on the particle physics community?

For so long now, the Higgs – quite rightly – has been the focus of this community. However, back in Munich [at the EGI Community Forum 2012, held in March], we established a strategy to give us a framework to bring the community together and direct it beyond the Higgs discovery: to be clear as to how we’re going to go out to new communities, how we’re going to broaden the appeal of the infrastructure, and how we’re going to continue to establish ourselves as a community. What we’re seeing here is those initiatives coming to fruition and now we’re getting a clearer, individual focus on what will carry us forward in the years to come.

Our focus, as much as possible, is on trying to grow our community into new research areas. The life sciences community has been very strong within the grid infrastructure over the last decade and what we’re doing now is very much trying to understand the switch from compute-orientation to data-orientation workloads. That broader focus on data means that we’re shifting our focus in terms of how we deliver the infrastructure out to those communities. If you’re running data services, you need a different form of infrastructure and infrastructure support than if you’re running compute-based services. The life sciences community has been engaging with the ELIXIR project to understand what their needs are for a European-wide infrastructure and to act as a provider of resources to them. They sit much closer to their user community and appreciate the nuances of different disciplines and sub-disciplines within that. We need to work with them, in order to best serve their users.

The Large Hadron Collider and the discovery of the Higgs has undoubtedly been the driving force behind grid computing for the last 10-15 years, there’s no denying that. It’s been a great way of pushing forward what we’ve been doing. It has dominated what we’ve done and what we’re now seeing is a rebalancing in priorities and in the type of infrastructure we deliver to cater for different communities. Obviously, the life sciences is one of these communities, but there are many others as well of course, all of which have this common challenge of high-throughput data analysis. That’s what we’re heading towards in the near future.

With the current global financial situation as it is, however, aren’t the really huge-scale science projects, which the grid infrastructure has typically served, now less likely to get the funding they require than they were perhaps just five or six years ago? As such, do you think that this “rebalancing” could also maybe be seen as part of a shift away from over-reliance on ‘big science’ projects?

I actually think it’s going to be the reverse of that. There are a lot more large science projects coming down the way and they’re going to be even more international, even more collaborative, even more distributed than ever, so the need for grid computing infrastructures to support the analysis is clear. Clearly, if you’re making a major instrumentation investment, you’re not going to be able to match that in terms of computing investment. You’re going to have to leverage the resources you have nationally, wherever you have them, and work to bring these resources together. That’s exactly the role we played behind CERN and the LHC and that’s exactly the role we’re going to be playing behind the new research infrastructures which are being set up in Europe. The economic climate has certainly caused a shift in focus  in terms of what new science projects do get the go ahead, but the need for continued investment and innovation is something that’s recognized by the European Commission and the national governments alike – that’s not going to change.

All in all then, would you say the forum has been a success?

Yes, definitely. These events are critical for bringing the community together, for resolving differences, and for understanding what we want to do going into the future. For me they’re priceless in terms of building the community. Having a good event, such as the one CESNET has delivered, is absolutely critical in terms of pushing us on into the next six months of activity.