Search and Solution Giants Unite in Clouds

By Tim Conneally, BetaNews

October 9, 2007, 5:58 PM

Sometimes, it takes an esoteric name to open the floodgates of interest onto an idea that's existed for a long time.

"Clouds" is a term describing software designed to be rich internet applications that balance the computing between thousands of processors working lightly, rather than fully taxing a single one. Many current cloud apps were once remanded to the user's machine, but can now be accessed online and with minimal demand on the user end.

In hopes of informing computer science students about the parallel-processing that takes place in the Cloud computing model, Google and IBM have announced that they would be teaming up to provide several colleges with large clusters of several hundred computers upon which to learn and test.

The servers, a combination of Google machines and IBM BladeCenter and System X machines, will run an open source implementation of Google's published computing infrastructure (MapReduce and GFS from Apache's Hadoop project).

Google and the University of Washington have developed a Creative Commons-licensed university curriculum that focuses on massively parallel computing techniques. Carnegie-Mellon University, MIT, Stanford, UC Berkeley, and University of Maryland are the schools participating in the program.

"This project combines IBM's historic strengths in scientific, business and secure-transaction computing with Google's complementary expertise in Web computing and massively scaled clusters," said Samuel J. Palmisano, chairman, president and chief executive officer, IBM. 

"We're aiming to train tomorrow's programmers to write software that can support a tidal wave of global Web growth and trillions of secure transactions every day."