CNET también está disponible en español.

Ir a español

Don't show this again

Christmas Gift Guide
Tech Industry

Commentary: A computing "space race"

Grid computing is much more than the latest fad. Stimulated by the Internet, computing is undergoing a profound transition into its third wave of evolution.

By Robert Batchelder, Gartner analyst

The idea of grid computing has been around since the days when computers were first networked. The academic community has long been fascinated with the potential to link together the world's computers to tackle massive computational and social problems, such as forecasting the weather and curing intractable diseases.

See news story:
Grid computing luring mainstream backers
For years, the practicality of attaining this vision has been more dream than reality, due in large part to the technical Tower of Babel that the computer industry has created with multiple operating systems, processor architectures, programming languages and networking protocols.

In combination, the technologies present themselves to application developers as dozens, if not hundreds, of different environments to support. Even Java, the Holy Grail of write-once, run-anywhere interoperability, collapses in the face of such a daunting task. Thus, grid computing has languished in the backwater of computing.

Grid computing may not be mainstream, but those devoted to it have persevered by addressing stubborn challenges in a very thoughtful way. The emergence of the Internet and the near ubiquity of Windows have accelerated their efforts.

Much of the work is represented in the Globus Project, an academic collaboration that has addressed and overcome many of the challenges presented by today's global, heterogeneous network of computers. Taking notice of this effort, the largest computer companies are integrating the project's results into industrial-strength grid computing environments.

These companies realize that grid computing will tie disparate desktop machines and servers together--and may even spur a demand for new machines. Given the importance placed on providing integrated hardware, software and services, Compaq Computer, IBM, Sun Microsystems and other companies recognize the wisdom in offering grid computing options to their customers.

However, grid computing is much more than the latest fad. Stimulated by the Internet, computing is undergoing a profound transition into its third wave of evolution. First-wave computing was defined by terminal/mainframe interaction. Second-wave computing was defined by client-server interaction. Third-wave computing will be defined by grid-peer interaction.

Today's hot "edge" technologies--such as peer-to-peer networking, content-delivery networks, instant messaging and Web services--are the first manifestations of a computational model, which will ultimately distribute application logic, transactions, collaboration and data horizontally across the globe.

Enterprises are preoccupied with vertically integrating their applications with the Internet and tools such as enterprise portals. As they succeed, they will begin to look for solutions that allow them to horizontally integrate their computing environments with those of their customers, suppliers and trading partners. Web services models--Sun One, WebSphere and .Net --will provide a substantial part of the fabric necessary for enabling interenterprise transactions.

Looking beyond the Web services model, similar capabilities will be needed to build the collaborative, computational and content networks of the future. IBM and the Globus Project's Open Grid Services Architecture is the first to take a comprehensive view of these disparate yet related requirements and attempt to rationalize them into a coherent computing landscape--an insightful move because each of the future networks share common problems. Solving problems in one type of network will benefit the other types. In that regard, Sam Palmisano, soon to be IBM's CEO, is precisely correct that Web services is a synergistic component within the overarching grid architecture.

No doubt, others will make similar proclamations to try to trump IBM and other competitors in this global computing architecture "space race." But as in the early days of the race to the moon, society is merely at the beginning of the decade when the Internet will transform itself into an ubiquitous fabric composed of grids (and peers).

In the short run, enterprises will have their hands full figuring how to integrate themselves horizontally with Web services tools that are not yet fully proven and trying to rationalize the accompanying transactional networks into a more seamless and generally useful computing infrastructure.

In the next few years, computational grids and then transactional (Web services) grids will provide the most useful benefits--which is what this buzz is all about.

(For a related commentary, on virtual computing, see gartner.com.)

Entire contents, Copyright © 2002 Gartner, Inc. All rights reserved. The information contained herein represents Gartner's initial commentary and analysis and has been obtained from sources believed to be reliable. Positions taken are subject to change as more information becomes available and further analysis is undertaken. Gartner disclaims all warranties as to the accuracy, completeness or adequacy of the information. Gartner shall have no liability for errors, omissions or inadequacies in the information contained herein or for interpretations thereof.