Commentary: Network appliances no replacement for PCs at home or office

Despite Oracle CEO Ellison's prediction of the decline of the PC, consumers still need to run their base applications locally, and the primary platform for doing that is the PC.

4 min read

Despite the Comdex pronouncement by Oracle CEO Larry Ellison of the imminent decline of the PC, both consumers and business users still need to run their base applications locally, and the primary platform for doing that remains the Windows PC.

Although more computing is moving to the Web,

See news story:
Ellison to Microsoft: Your software is too complex
the "network computing" approach still faces an uncertain outlook. Network computing (via thin appliances) may have a stronger future when networks are more available--for instance, when dependable, high-speed wireless networking is commonly available and inexpensive.

The issue then will be the cost of accessing software on the grid, versus owning it on your PC. At least then we could have the debate. But our research indicates that this is probably three or more years away.

Some niches do exist where network appliances may make sense. For instance, some portion of the public does not want to deal with any level of technical complexity. They might be more open to using very simple, small, network-only devices in their homes.

But consumers are more likely to use network appliances to supplement PCs instead of replacing them. This is analogous to consumers purchasing a stove, a microwave oven, a toaster oven and a coffee maker. You could do it all with a stove, but for convenience you add the other targeted appliances to the mix as finances allow. If you could only afford one, then it would be the stove.

For businesses, network appliances shift the complexity of system setup and maintenance from the desktop to central servers, which are more accessible to IT support workers.

There are some niches where the point is not to save money in the initial purchase but in ongoing support. Network appliances and "locked-down" PCs are so simple that they need very little support, and consumers cannot add things that create problems.

But advances in desktop management are providing similar centralization of administration to traditional PC based desktops, thereby blunting some of the benefit seen with network devices.

And while network appliances may sound less expensive to buy than PCs, their predecessors--X-terminals and network computers--were not cheaper. And they proved inflexible and unable to upgrade as technology changed--specifically, with the introduction of the Internet.

IT departments that bought into network computing in the past are now facing some limitations because they cannot be upgraded. PCs are inherently more flexible and expandable, and that makes them a better investment in the long run than network computing.

The vast majority of users constantly need more local computing power to run their base applications, independent of network availability issues. One of the main reasons why companies discarded "dumb" terminal computing was that every network slowdown hampered the productivity of every person in the office.

Today office productivity is even more closely tied to computing, and workers need to be able to be productive even when the network is down, or when they are unable to connect at all--for instance, while traveling.

Long-term, client-side processing power needs to be leveraged to enhance interaction and provide the next level of user productivity gains.

While the "application logic" may rest on servers, the "presentation logic" will continue to evolve and take on greater importance. Technologies such as speech, rich media and natural language will need to be handled at the device, requiring ever more powerful clients than we see today. Appliances today simply don't handle those technologies well.

Locked-down computers, which prevent users from adding games and other personal software and which force them to store their data on central servers, where it can be backed up regularly, are often better solutions for gaining more control over a business computing environment.

Although there will be more network-connected non-PC devices (such as cell phones and PDAs), even these have significant local computing power (the typical $150 MP3 player has 48MB of flash memory, a basic processor and a simple operating system), and they will complement desktop PCs rather than replace them.

Over time, however, we expect to see some blurring of the line between appliances and traditional PCs, as operating system code becomes more modular and user platform management more automated. All devices will exhibit both thin and thick characteristics, in varying degrees.

Businesses may want to consider network appliances (primarily sub-$1,000 locked-down PCs) for specific niche needs. Consumers who are averse to dealing with any kind of complexity may be willing to trade the flexibility of a PC for the simplicity of a network device.

But both groups should keep in mind that network appliances deliver much less capability for about the same purchase price. They cannot be upgraded as network technology--which is in a high state of flux. They cannot operate independently of the network.

For consumers, the lifetime cost of ownership of these devices will depend on the cost of Internet and resource access. If that rises over time, their computing costs will rise with it, and if they have vital personal information on a Web server, they will have little choice but to pay those costs.

Oracle, of course, has an ax to grind, as Ellison wants to knock Microsoft off its pedestal. But the emphasis on PC replacement is entirely the wrong focus. What we are seeing is an expanding role of end-user computing environments, where various types of devices pervasively expose services to consumers, employees, partners and suppliers.

The PC remains a foundation for the foreseeable future. It will be supplemented by other devices, such as network appliances, which act as touch points around user work and lifestyle activities. META Group analysts Dale Kutnick, Val Sribar, William Zachmann, Steve Kleynhans, and Mike Gotta contributed to this article.

Entire contents, Copyright © 2000 Meta Group, Inc. All rights reserved.