IBM creates autonomic computing unit

CIOL Bureau
New Update

By Caroline Humer

NEW YORK: IBM Corp. this week will unveil plans to broaden its efforts to develop computer systems that not only manage but also fix themselves when they break. Armonk, New York-based IBM said it is creating a separate division, which it calls Autonomic Computing, which will integrate its efforts to create computer systems that need less human intervention. As part of the software division, the unit also will work on standardizing those efforts to work with software programs from other companies.

The move is part of a broader goal the computer services, hardware and software company has of bringing computing to customers on demand, said Nick Donofrio, IBM senior vice president of technology and manufacturing. Autonomic computing moves the complexity of computing away from the user, he said.

"We've been hyping the pizazz out of this thing for 12 months. It is now time ... to start to help people understand these are the directions we're going in, these are the standards," Donofrio said. IBM launched an autonomic computing initiative in its research division in the spring of 2001, aiming to bring some of the traditional characteristics of mainframe computing to other, less expensive systems.

For instance, IBM said it includes autonomic capabilities in its Tivoli management software and DB2 database software and is adding the features to its "Shark" storage system. This week, the heads of IBM's computer, software and data storage businesses will discuss the plan when they meet with chief information officers for three days in Desert Springs, California.

With technology budgets expected to increase only a few percentage points in 2002 after having fallen in 2001, IBM said it aims to save customers money by decreasing the number of people needed to manage and fix complex computer systems.

Alan Ganek, an IBM veteran who previously was in charge of strategy for IBM Research, will head the unit. Part of his job will be getting customers more comfortable with letting computers come up with advice on how they can be used better, and then enabling those systems to take that advice without the go-ahead from technology workers, he said.

"That's absolutely an issue and that's why our customers want an evolution, not a revolution," Ganek said in an interview. Analysts agree.

"People will really need some time before they really trust this technology to do the right thing. I think there are some impediments that are not so much about technology -- they are more about psychology," said Mike Gilpin, an analyst for Giga Information Group. "But these developments will be quite welcome if they really do work. That remains to be seen."

© Reuters