Ian Foster

Director of the Computation Institute at Argonne National Laboratory

Q: Let's start with the shameless plug part: What are you working on right now, why should your average user or developer care, and why is it super cool and challenging?

A: Let me mention two projects that I'm particularly excited about. The first is CIM-EARTH, the Community Integrated Model of Economic and Research Trajectories for Humankind. This collaboration among economists, geophysicists, mathematicians, and computer scientists aims to produce better understanding of the complex interactions between human economic activities and environmental change, based on better and more open data and models. We've recently been funded by the NSF Decision Making Under Uncertainty program to establish a new Center for Robust Decision Making on Climate and Energy Policy, which will advance this goal.

The second is an as-yet-unnamed project that is using software-as-a-service methods to deliver process automation and application software to researchers. Our ambition is to move much of the complex IT out of the researcher's lab and into managed services. I believe this change of focus will be transformative. We've started with the deceptively simple problem of data movement. Users can hand off data replication and mirroring tasks to our Globus.org data service, which then manages all the complexities of establishing, managing, and optimizing a transfer. The user installs no software. We have an ambitious program of work defined beyond that, but data transfer is a nice first step.

Q: Are you well-known in the field? Are conferences the place where everyone knows your name? Are colleagues more likely to know of you than to know you personally?

A: Google scholar says that my papers have more than 50,000 citations. That doesn't mean 50,000 readers (cut and paste is so effective), but it does suggest that many people know of my work.

Q: What work of yours has recently received attention within the field or media?

A: A recent article in IEEE Computer (Parallel Scripting for Applications at the Petascale and Beyond) makes the case for parallel scripting as a programming model for ultra-scale computers. I think that the community's focus on only scaling SPMD computations is misguided, and so I hope this article receives attention.

Q: Are you getting jaded about interviews with press? When was the last time you or your work was featured?

A: I always find it interesting to talk to the press, because they are concerned with explaining technical concepts to a general audience, an important task that I'd like to be better at myself.

The media likes to talk about cloud computing at present, so that has been the subject of some recent conversations. Sometimes I quote Miron Livny, who quips that he has been doing cloud computing since before it was called grid computing!

Q: How trendy is the topic you are currently working on? Are all the cool kids doing it, or will they be in another year?

A: I hope that the two areas that I listed at the beginning become trendy. We urgently need better and more open models of energy and climate economics, but they've been a fringe topic since the 1970s. The software-as-a-service platform and app store that we are developing is perhaps a bit ahead of where most people are with "cloud", which is outsourcing computing and storage. In both cases, we hope that some cool kids will want to participate in what we are doing.

Q: Has anyone ever acted like they were your Biggest Fan Ever, or said something like, "I love your work?" Tell us a story, or toss us the most fan-ish comment you recall receiving.

A: I'm sure people have said fan-ish things to me, but being a modest New Zealander I immediately forget such conversations.

Q: What is the most controversial aspect of you or your work? Come on, you can tell us!

A: I don't think I have ever said or done anything controversial. That's a bad sign, and I hereby resolve to correct this unfortunate situation.

Q: Do you have people? Rock stars always have people who talk to other peoples' people.

A: I have two people who are always ready to speak truth to power: my 13-year-old son, who I desperately call "shorty" because he will be taller than me within months, and my 11-year daughter, who tells me that I am not a real runner because I haven't qualified for the Boston marathon. They're both available to talk to other peoples' people at any time, although I may not approve of what they say!

David Anderson

Principal Investigator, BOINC

Q: Let's start with the shameless plug part: What are you working on right now, why should your average user or developer care, and why is it super cool and challenging? Please try to respond in one paragraph, using as little jargon as possible.

A: Right now, my group is working, together with Ben Segal from CERN, on the "volunteer cloud", which means running volunteer computing applications in virtual machines. This will make it easier for scientists to use volunteer computing by eliminating some of the headaches cause by multiple platforms like Windows, Mac OS X, Linux, and variants. We're working on other stuff which is challenging but not cool.

Q: Are you well-known in the field? Are conferences the place where everyone knows your name? Are colleagues more likely to know of you than to know you personally?

A: I personally focus on developing software rather than on writing papers or attending conferences. As a result I'm essentially unknown in the computer science world.

Q: What work of yours has recently received attention within the field or media?

A: Einstein@home recently discovered an unusual pulsar and announced it in Science, generating some media attention. Bruce Allen, who runs Einstein@home and directs the Max Planck Institute in Hannover, was kind enough to include me as a coauthor on this paper.

Q: How trendy is the topic you are currently working on? Are all the cool kids doing it, or will they be in another year?

A: Volunteer computing isn't trendy, but it's not going away: a billion GPUs are the world's biggest computing resource for the foreseeable future.

I'm also involved in "scientific crowdsourcing", which is like volunteer computing but using people instead of computers. Scientific crowdsourcing is becoming trendy, but we're still looking for a killer app.

Q: Has anyone ever acted like they were your Biggest Fan Ever, or said something like, "I love your work?" Tell us a story, or toss us the most fan-ish comment you recall receiving.

A: I have a few fans among die-hard BOINC and SETI@home volunteers. Most of these fans resemble Comic Book Guy from The Simpsons.

Q: What is the most controversial aspect of you or your work? Come on, you can tell us!

A: Many people in the HPC world still can't wrap their heads around the idea that you can do real, hard-core scientific computing on home PCs. I think there are deep issues involving ownership and control - volunteer computing provides neither.

Q: Do you have people? Rock stars always have people who talk to other peoples' people.

A: No; I answer my own phone.

Mark McAndrew

The Charity Engine

Q: Let's start with the shameless plug part: What are you working on right now, why should your average user or developer care, and why is it super cool and challenging? Please try to respond in one paragraph, using as little jargon as possible.

A: With a bit of luck, we're about to create the most powerful and energy-efficient computing platform on Earth. It's called The Charity Engine, a volunteer network with two crucial differences. Firstly, it's going to be big - really big. Secondly, any ethical research will be able to use it on a simple pay-as-you-go basis, because our volunteers will not be choosing between science projects. It will be a worldwide computer for hire.

Q: Are you well-known in the field? Are conferences the place where everyone knows your name? Are colleagues more likely to know of you than to know you personally?

A: You'd have to ask them. All I can say is that we've had to reorder our business cards. Twice.

Q: What work of yours has recently received attention within the field or media?

A: That'll be The Charity Engine. We like to focus on just one world-changing project at a time.

Q: Are you getting jaded about interviews with press? When was the last time you or your work was featured?

A: By astonishing co-incidence, this is our 100th* press interview - so you would be forgiven in thinking so. However, The Charity Engine is always a pleasure to discuss. (*Disclaimer: may refer to binary notation.)

Q: How trendy is the topic you are currently working on? Are all the cool kids doing it, or will they be in another year?

A: About as trendy as a Dungeons & Dragons convention in 1985. Right now, we're all nerds - but the cool kids will certainly be doing this next year. They'll be us.

Q: Has anyone ever acted like they were your Biggest Fan Ever, or said something like, "I love your work?" Tell us a story, or toss us the most fan-ish comment you recall receiving.

A: At the moment of writing, The Charity Engine has exactly 94,696 fans from at least 120 countries (see www.tinyurl.com/charityengine). Not bad considering we haven't launched yet. A particularly nice comment involved the promise of a free laptop, but we had to respectfully decline on the grounds of competence.

Q: What is the most controversial aspect of you or your work? Come on, you can tell us!

A: Explaining that a zettascale computer already exists: the Internet itself. We just need to use it properly.

Q: Do you have people? Rock stars always have people who talk to other peoples' people.

A: Naturally, which is why my PA is answering these questions on my behalf. That said, keeping to a strict start-up budget means that the CEO and his Personal Assistant are currently sharing the same responsibilities and, indeed, physical form.

Thank you for your kind attention. If your organisation would benefit from ultra-cheap, ultra-large scale distributed computing resources, you can contact The Charity Engine using info (at) charity-engine (dot org). The worldwide computer goes live in 2011.