By Allison Proffitt
July 29, 2010 | ‘Most people have a very narrow definition of the cloud. Most people, when they think of the cloud, think, ‘How can I get my computer over the internet?’ They think of the Amazon EC2 model. But that’s just the beginning.”
So says Chris Whitney, director of HP Labs Singapore, who believes the cloud story is based on “everything as a service” and that “Cloud 3.0” is not far away. The company’s new lab in Singapore will be home to some of the research and “deep thinking” around the evolution of cloud computing.
The lab in Singapore is the newest of HP Lab’s seven sites, employing about 500 people worldwide. The Palo Alto lab is the headquarters, and Singapore joins labs in Bristol, UK; Haifa, Israel; St. Petersburg, Russia; Bangalore, India; and Beijing, China. Those seven labs focus on eight general themes across HP’s business, including Cloud computing. The Singapore lab will be “predominately cloud” focused.
With labs in China and India, Whitney says that HP’s Singapore lab will be the hub for Asia Pacific, including Australia, outside those two countries. Singapore’s Economic Development Board and A-STAR—as usual—have made setting up shop in Singapore easy on HP. The lab is located in Fusionopolis, Singapore’s recently-built R&D hub for Infocommunication Technology, Media and Physical Sciences & Engineering.
Managing the Cloud Stack
HP began developing its internal Cloud computing resources more than two years ago, says Whitney. “I was actually looking after [an HP Lab] in Palo Alto, and I’d been working very closely with [the HP] lab in the UK. We got together and we invented this program called Open Cirrus [with partners Intel and Yahoo!]. Serious—it’s supposed to be a play off the cirrus cloud and getting serious about the Cloud.
“We were watching everything that was happening for example with Amazon with things like EC2, looking at what Google was doing with Google Apps, we were starting to understand what Microsoft was doing with things like Azure,” Whitney says.
After looking at the landscape, HP chose a multidirectional attack. Some HP teams are taking a top-down approach: looking at how businesses actually use the cloud. Other teams are taking a bottom-up approach: addressing concerns around guaranteed security and guaranteed performance. “It’s really trying to make the whole notion of infrastructure as a service business-ready,” Whitney explains.
Part of the Singapore lab’s mission will be what’s left: the meat in the middle. In the industry it’s called platform-as-a-service at the moment,” Whitney explains. “We believe in the future everyone will be walking around with devices like these [iPhones] accessing services generation-wide. You can do anything you want from getting CPU power, running services, running applications on these things, and we were thinking, well, what does that really need? If you think of cloud as the next evolution of the services and IT industry, and if you think about this ‘everything-as-a-service’ vision where everyone’s using these services, what’s needed in the middle? Either services or software stacks or technology to do that bridging.”
HP’s goal is to enable cloud business, to allow individuals, companies, and governments with a vision to offer services on the cloud to their customers. In Singapore, HP is working with the government to build “government clouds” and is also focusing on education and media. Working with the Malaysian Institute of Microelectronic Systems (MIMOS), HP researchers are even looking at ways that cloud services can enable better agriculture.
“We want people to be able to concentrate on their imaginations, not get put off on, ‘Well, I don’t really know how to build this application, therefore I can’t really play in this landscape.’ We want people to get out there and build their application, get out there and make money and take part in that ecosystem without them having to be a deep Java or Microsoft tech type of people.”
Wheelbarrow Full of Chips
The second directive for the Singapore lab is to look at how to build a cloud data center. Whitney is candid about Singapore’s challenges in this area. “Land is expensive here. There are no natural resources; it’s 35 degrees C and humid on a good day. If we can make data centers cost-effective here, we can make them cost effective anywhere.”
A new generation of computing calls for new ideas around data centers, Whitney believes. “I call it the postmodern data center project,” he says. “If I went to the car park with a wheelbarrow of chips and dumped them down there and said, ‘Build me a cloud data center,’ what would you build me? If you think of the two ends of a data center, one end is the chip… because the data center is really all about processing, and the other end is a concrete shell to keep the rain out.”
With only those two limitations, “Should we look for a square building? If we built, for example, a wind-powered data center, could the data center be in the stem of the turbine?”
Cloud 3.0 is “a lot of technologies that we’ve been putting in the data centers that allow us to do, for example, data sharing, to be able to do sharing of CPUs, people can actually get together and do mashups very quickly,” says Whitney. “It’s nothing different. It’s just an evolution of the thinking that took place on the internet, on voice-over-IP, on email, and on the Web. It’s just the next evolution of where these things are going.”
This article also appeared in the July-August 2010 issue of Bio-IT World Magazine. Subscriptions are free for qualifying individuals. Apply today.