Bloomberg Businessweek
By Aaron Ricadela
Researchers seeking smarter ways to tackle the most complicated computing tasks think they've found the answer in a cloud—though not the kind that wafts across the sky as masses of condensed water droplets and frozen crystals. Instead, they're turning to something called cloud computing, which aims to deliver supercomputing power over the Internet.
IBM is the most recent company to announce plans to tap cloud computing technologies. On Nov. 15, IBM executives in Shanghai unveiled a system, dubbed Blue Cloud, that will let banks and other customers distribute their programs across large numbers of machines to deliver faster, more sophisticated data analysis. The first Blue Cloud products are due in the spring of 2008.
Supercomputing for the Rest of Us
Two top Internet companies recently announced similar projects. Yahoo! on Nov. 12 said Carnegie Mellon University, and eventually other schools, will use a 4,000-processor computer housed at the Web company to conduct software research. And Google, the steward of what's effectively one of the world's largest supercomputers used to power its search engine, in October said it would make hundreds of processors in its data centers available to schools including the University of Washington, Stanford University, and MIT to help teach high-performance computing programming techniques.
"All of these are examples of the frenzy around cloud computing," says Dan Reed, a longtime supercomputing researcher who will start work as Microsoft's (MSFT) director of Scalable & Multicore Computing on Dec. 3. Fueling that frenzy, says Reed, is the proliferation of high-speed Internet connections, cheaper and more powerful chips and disk drives, and the development of data centers that house hundreds or thousands of computers to quickly serve sophisticated software to legions of users. "None of this would have been possible a decade ago," he adds.
Herewith, a primer on how companies—and consumers—might harness cloud computing's power:
How does cloud computing work?
Supercomputers today are used mainly by the military, government intelligence agencies, universities and research labs, and large companies to tackle enormously complex calculations for such tasks as simulating nuclear explosions, predicting climate change, designing airplanes, and analyzing which proteins in the body are likely to bind with potential new drugs. Cloud computing aims to apply that kind of power—measured in the tens of trillions of computations per second—to problems like analyzing risk in financial portfolios, delivering personalized medical information, even powering immersive computer games, in a way that users can tap through the Web. It does that by networking large groups of servers that often use low-cost consumer PC technology, with specialized connections to spread data-processing chores across them. By contrast, the newest and most powerful desktop PCs process only about 3 billion computations a second.
Which companies are at the forefront of cloud computing?
Google's search engine and productivity applications are among the early products of efforts to locate processing power on vast banks of computer servers, rather than on desktop PCs. Microsoft has released online software called Windows Live for photo-sharing, file storage, and other applications served from new data centers. Yahoo has taken similar steps. IBM has devoted 200 researchers to its cloud computing project. And Amazon.com recently broadened access for software developers to its "Elastic Compute Cloud" service, which lets small software companies pay for processing power streamed from Amazon's data centers.
What's the market opportunity for this technology?
While estimates are hard to find, the potential uses are widespread. Rather than serve a relatively small group of highly skilled users, cloud computing aims to make supercomputing available to the masses. Reed, who's moving to Microsoft from the University of North Carolina, says the technology could be used to analyze conversations at meetings, then anticipate what data workers might need to view next, for example. Google, Microsoft, and others are also building online services designed to give consumers greater access to information to help manage their health care.
What are the biggest challenges these companies face?
The technical standards for connecting the various computer systems and pieces of software needed to make cloud computing work still aren't completely defined. That could slow progress on new products. U.S. broadband penetration still lags that of many countries in Europe and Asia, and without high-speed connections—especially wireless ones—cloud computing services won't be widely accessible. And storing large amounts of data about users' identity and preferences is likely to raise new concerns about privacy protection.
Haven't we heard about efforts like this before?
Every decade or so, the computer industry's pendulum swings between a preference for software that's centrally located and programs that instead reside on a user's personal machine. It's always a balancing act, but today's combination of high-speed networks, sophisticated PC graphics processors, and fast, inexpensive servers and disk storage has tilted engineers toward housing more computing in data centers. In the earlier part of this decade, researchers espoused a similar, centralized approach called "grid computing." But cloud computing projects are more powerful and crash-proof than grid systems developed even in recent years.
