November 15, 2004
GRIDtoday editor Derrick Harris was able to speak with United Devices president Ed Hubbard about his company's role in developing the World Community Grid, among other topics, including Grid's evolution since SETI@home, and why skeptics need not worry about the security of public Grids.
GRIDtoday: To start with, I am wondering why IBM selected United Devices to power the World Community Grid.To start with, I am wondering why IBM selected United Devices to power the World Community Grid.
ED HUBBARD: Well, I think IBM looked around at the available options. There's a few key ingredients to build one of these Grids. One is, "can it scale?" because they want to build a large Grid. We've used our infrastructure for the last four years to run a public Grid of our own, which resides at Grid.org, and that's about 3 million nodes today, so they were pretty pleased with the scalability picture. Of course, we built the thing so that it used DB2, which was also a huge plus for IBM -- they love seeing their database used.Well, I think IBM looked around at the available options. There's a few key ingredients to build one of these Grids. One is, "can it scale?" because they want to build a large Grid. We've used our infrastructure for the last four years to run a public Grid of our own, which resides at Grid.org, and that's about 3 million nodes today, so they were pretty pleased with the scalability picture. Of course, we built the thing so that it used DB2, which was also a huge plus for IBM -- they love seeing their database used.
So, scalability was one, security was another. There's a lot of ... distributed computing platforms out there, like folding@home, SETI and a bunch of others, but those guys, while they're pretty secure, they didn't really spend as much time as we did on the security model. That was another thing that IBM reviewed very carefully, I think, and they liked what they saw.
I think the third point, and I would encourage you to talk to IBM about this, I think one of key ingredients was also the fact that our infrastructure can be used, and has been used, that's how we make our money, in corporations. The same infrastructure that runs on the public Grid is the same infrastructure (software) that we sell to companies around the world. Using a commercial product like this on the public Grid, especially when it can be used inside of companies, I think was a big plus.
Gt: Alright. Now, am I correct in my belief that Grid.org and World Community Grid will be working as one throughout the duration of the Human Proteome Folding Project, and then are going to be working on separate projects afterward?
HUBBARD: That's exactly right. They will be working together on this first project of IBM's while IBM gets its Grid up and running, and then the World Community Grid will be a standalone entity that IBM controls and runs, and someday may turn into a real non-profit altogether. Grid.org will continue to be ours and we'll run it separately and do other projects on it.That's exactly right. They will be working together on this first project of IBM's while IBM gets its Grid up and running, and then the World Community Grid will be a standalone entity that IBM controls and runs, and someday may turn into a real non-profit altogether. Grid.org will continue to be ours and we'll run it separately and do other projects on it.
Gt: What is your forecast for the success of the World Community Grid, in particular this first project that you're a part of, based on your successes in the past with Grid.org's projects:
HUBBARD: Well, I think IBM will be very successful. Whenever a large company like this gets behind one of these projects, you get a lot of focus and a lot of people want to join. That's basically why we built Grid.org. We've run three other projects out there in the past: one around cancer, one around anthrax toxin, and one around looking for an anti-viral for smallpox. We've worked with Oxford University in all those cases, but the critical piece of those projects was getting sponsorship from large technology companies, like IBM, Intel and Microsoft had all been involved in the past.
These things take on a life of their own, and before you know it you have millions of machines out there. It's an interesting responsibility you have to keep those machines fed with something the user base likes to see.
Gt: You mentioned SETI@home before, how do you think the public Grid concept has evolved from the launch of SETI@home to the current World Community Grid launch?
HUBBARD: Well, I think there's couple of key differences and evolutionary points. One is that the SETI@home guys, and all the other @home folks, were really: one, they were rolling their own infrastructure; and two, they were really focused on their particular project. So downloading SETI@home, for instance, meant that you were just going to run an analysis program for radio telescope data, and that's it.
When you download an infrastructure like ours on Grid.org or IBM's at World Community Grid, it's a general purpose platform that can run multiple projects simultaneously. IBM has already said publicly that they want to run other projects, even while the folding project is going on they may bring other projects online and run them on the same infrastructure, and let users have a choice in what they run, which is always key. I think that's been one big difference between the SETI@home model and this evolving model.
It's really interesting, it's sort of like the old operating systems. When you bought a word processing system, it had an OS, and it was a word processor, and that's all it did. Then came the current of innovation of, "Hey, listen the operating system and the applications can be separate, and I can run lots of different applications on top of one operating system." That became a pretty standard way to think about applications and operating systems, and I think the analogy is a pretty good one here, as well.
Gt: Keeping along that same line, a Columbus, Ohio, man was fired last month for running SETI@home on his work computer, and I was wondering what you might say to skeptics, like the man's employer, perhaps, who are concerned about security or performance issues stemming from running Grid software on a PC?
HUBBARD: Well, just like IBM has done, and we've done in the past, we always encourage people to only run it on machines that they have authorization to run it on. Sometimes people don't really think about that before they install software. This software is no different than installing any other software that they might come across on the Internet -- some other screensaver or, remember in the old Pointcast days when people would download that and get in trouble. There's really no difference here, it just happens to be a phenomenon that is pretty widespread.
Do I think that the people who own those machines are being rational? No, I don't. I think it's pretty irrational to fire someone for running one of these projects because, in general, they're very secure. They're only reporting results back to a single server out there, and you know exactly who wrote the software. It's either some folks at a university or, in our case, it's United Devices' and IBM's corporate logo on the line, basically, and they don't mess around with their logo. They will not put their brand in jeopardy. So, I think it's irrational and over time, with any new technology there's resistance and misunderstanding, and then over time it becomes much more widely accepted and understood and nobody's afraid of it anymore, but it still happens from time to time.
Gt: The World Community Grid is just one of a whole rash of Grid projects that have been announced, a lot of states are running statewide Grids, for example, and I was wondering what you see in the future for large-scale public Grids, and how much the general public stands to benefit from the technology?
HUBBARD: Well, a few answers come to mind. One, I think we've been running this kind of model at Grid.org for the last three and a half/four years, and SETI@home has been around for a long time, but this is the first time that a really large, leading technology company has come out and said, "This is something that's important and we're going to be involved. Not in just a sponsorship, but we want to run one of these. We think it's going to make a difference to our company and to the researchers involved with the projects we run." So, this is the first sort of major stamp of approval endorsement, if you will, from a large technology company like this, which I find fascinating. That's just on the public side, it has nothing to do with the private Grids.
We have probably 50 customers around the world that have set up and run private Grids on our technology, but they don't get the notoriety or the attention that these public ones do by definition.
Where does this go after this? I think it is interesting to see these states start to think about putting their own Grids together. For us, we're just in the business of helping people build Grids, whether they're behind the firewall or on the public IP network like this, so we're quite happy to help anybody who wants to build a Grid, just like IBM. We have had some folks from various states around the country contact us today because they saw this happen, and they said, "You know what, we've been thinking about doing something like this in our state and we want to talk to you." I think that only happens when you get an endorsement like this from IBM.
Gt: The Grid has been getting a lot of press from the announcement today and I was wondering, although you kind of answered this already, how does this attention affect United Devices in regard to gaining customers for your products, enterprise Grids specifically?
HUBBARD: It's just a fantastic endorsement for us. This is one of those that you dream about when you start a company ... "Maybe, someday, IBM take the technology and do something interesting with it, and then that opens the floodgates." We've done well on our own, but I think this will change things a little bit for us.
Gt: Well, Ed, that's all the questions I have. Is there anything else you would like to add in regard to this project or United Devices in general?
HUBBARD: Outside of United Devices, we think this is a really interesting project from a scientific perspective. The Human Genome Project was really important and it got us the formula for being a human being, but it didn't really tell us how to bake the cake, or how the cake is baked. This project is going to attempt to go through the human genome, as well as a lot of other genotypes, and figure out what makes these things tick from a protein standpoint. It's going to be amazing to see the results of this when they come out. ISB (Institute for Systems Biology) is going to make the database of protein structures public so everybody in the world will be able to get access to these. Basically, were going to take the whole industry from knowing perhaps 30 percent of the proteins in the human being, and a lot of other organisms, to, in the next year, having at least a software-based prediction of what all the proteins are in everything we run on this, which is going to be a lot of different organisms.
So, from all of recorded history to yesterday, we know just 30 percent of the proteins in a human, and over the next 12 months we're going to get the other 70 percent of the proteins, plus a whole bunch of other organisms, completely run through as far as protein structure predictions go. We're just going to have this massive data that all of a sudden, just in a course of 12 months, there's just going to be an explosion of knowledge in what makes everybody tick. I think it's just incredible.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 10, 2013 |
Australian visual effects company, Animal Logic, is considering a move to the public cloud.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.