IBM CEO: Replacement cycle a dinosaur

The technology replacement cycle--the frenzied buying spree that the tech industry has come to count on every few years--is being driven to the tar pits by on-demand computing, said IBM CEO Sam Palmisano.

Speaking at a meeting with security analysts in Boston, Palmisano postulated that large corporate customers are no longer interested in investing in PCs or the latest servers.

Instead, these companies want to obtain software, consulting services and back-end hardware in a holistic manner to help them automate business processes and make their overall operations more efficient--the very activity that the Armonk, N.Y.-based giant has been promoting.

"It has been product-centric, gizmo-driven growth, but that is not going to happen again," Palmisano said. "This industry continues to be a growth industry, because it solves a very important problem: productivity."

Additionally, he said, large companies are once again funding internal technology projects.

"We really do see more (funding) now than a year ago, now that projects are going forward. Things have stabilized," he said.

Since taking over IBM last year, Palmisano has promoted a strategy called on-demand computing, under which large companies will provide encompassing solutions to existing business problems through technology.

How these solutions are implemented, he said, varies by the customer and industry in question; some corporations are looking for complete technological retrofits, while others will merely want to rent extra computing capacity for peak periods.

As a consequence of these trends, consulting and software for weaving together these organic systems will become a larger part of the profit pool in the industry. By 2005, these categories will account for 65 percent of the technology industry's profits, up from 46 percent today. Servers, storage systems, PCs and components, meanwhile, will decline from 54 percent of the profit pool to 35 percent.

Palmisano said that IBM is one of a very few companies that can provide all of those pieces. It has an extensive services and consulting group, participates in all of the relevant product markets, and can boast more depth in its research department than almost anyone else, he said.

Research, he said, "is why we are gaining all the share. It shows up in our servers. It shows up in our software. It shows up in our storage systems."

IBM has set up 17 separate practice groups to focus on solving business problems with technology in different industries. The life sciences group, for instance, focuses on using technology to make drug research more efficient. The group has gone from "zero to a half billion dollars," Palmisano said.

"The more business process insight we can bring to a problem, the more margin opportunity we have," he said.

Some hurdles ahead
Nonetheless, he acknowledged that all is not sunny in the services-centric future. In recent months, IBM has had to renegotiate consulting and services contracts, reducing the overall value of the deals. The company has also found itself saddled with excess capacity in services.

Furthermore, companies are not fully funding multiyear, major technology projects in the same manner that they used to. Rather than commit to a complete budget up-front, companies will fund a project for six months and, if it goes satisfactorily, fund another six months of work.

Invite Michael Kanellos into your in-box Senior department editor Michael Kanellos scrutinizes the hardware industry in a weekly column that ranges from chips to servers and other critical business systems. Enterprise Hardware every Wednesday.

IBM, Palmisano added, won't be getting out of the PC or chip business anytime soon, something some analysts have recommended for years. Large customers still need PCs, he pointed out. The company's PC and Intel-based server operations, moreover, are crucial laboratories for learning how to remove costs out of other operations.

Likewise, the microelectronics business is necessary to enhance the services and storage business. "We are in the business fundamentally to provide differentiation to our storage and server business," he said.

IBM will not become a large-scale foundry and manufacture chips for a wide variety of customers, Palmisano said. However, it will sign chip development deals with a dozen or so companies such as Sony that will use IBM's chip facilities.

"We are not going to become the next TSMC (Taiwan Semiconductor Manufacturing Co.). We are not going to become the next foundry or the next Intel," the CEO said.

In contrast to his predecessor, the often opaque Lou Gerstner, the jocular Palmisano took the opportunity to throw a few jabs at competitors, although largely anonymously.

After Hewlett-Packard acquired Compaq Computer, the combined company became for a time the largest overall server company. "We had the opportunity to buy a PC company in the past," but chose not to, he said.

People crowed that the combined company would be No. 1 in the server market, he said. "They were No. 1 for about four weeks. We're back to being No. 1."

Without naming Intel, he noted that a large semiconductor company recently announced that its top-end chip contained a flaw and could not be run at 1GHz, a problem that Intel noted with its Itanium chip.

"We've got three times the performance at 70 percent of the cost," Palmisano said.

IBM is No. 2 in storage, according to market research surveys, but next year the company should be No. 1, he predicted, a swipe at rival EMC.

And, as a veiled message to Microsoft, Palmisano touted corporate leaders' embrace of Linux, the open-source operating system. It's one of the topics that comes up most frequently in meetings, he said.

"When CEOs are talking about Linux, the world has changed," he said.