The skinny on PC vs. NC.

Is the PC an essential tool for corporate productivity or just a hairball on the desktop? Is the NC a sleek thin-client or a highly limited dumb terminal? The battle of the vendors for desktop IT budget dollars has been raging for years, but no clear winners have yet emerged. On the one side are the McNealys and the Ellisons, extolling the virtues of network-centric computing; on the other are the Gateses and the Dells, dismissing the NC as a passing fad. In the middle are chief executives, wondering whom to trust. As a CEO, you've heard the hype, the promises - and the mud-slinging. Now, read what the experts have to say.

The Great Debate

Pro PC

Michael Dell Dell Computers

"What I think most customers are looking for is a consistent architecture. NCs have some pretty serious flaws. For example, a lot of our customers are mobile users. They go on airplanes and want to take their computers with them. Well, if you have an NC and go on an airplane, you don't have anything. There's really no mobile architecture for the NC. So I think the managed PC is the right way to evolve the architecture."

Bill Gates Microsoft

"PCs can be networked, and most of our software is related to networks. People underestimated the PC phenomenon. Prices have come down. You won't recognize the PC five years from now because of speech, flat screen, higher performance, but it will still run the applications we have today. We believe in compatible devices... The cost of hardware is not the issue. You can get [a desktop device] where the hardware costs less, but you can't run when the network is down. NC stands for Not Compatible."

Pro NC

Larry Wilson Oracle

"We will never have an Information Age if we rely on the PC. PC penetration in the U.S. is only about 30 percent, leaving 70 percent of America disenfranchised...The rate of adoption of PCs is going down, not up. The PC is too complex to be a sensible application for general use in the home or office."

Scott McNealy Sun Microsystems 

"I think most mere mortals don't deserve a telephone switch and they don't deserve a personal mainframe on their desk or in their lap or in their luggage. I have yet to see any investment analysis on a PC show anything other than a negative return - unless you count employee satisfaction and thrills as psychic income and part of the ROI... PC-compatible? That's an oxymoron."

on the debate

John McCarthy Group Director Forrester Research 

We have statistics that show unequivocally that NC doesn't solve the actual problem at hand. People are still running applications - whether on desktop or server - and that's what they've got questions about. The real problem is that companies never want to invest in the soft dollar training and support. Just like they didn't want to invest in factories in the '60s and '70s - and that came back to bite them in the ass.

The dirty truth is that Scott McNealy's IT group has written a custom set of applications to run on his NC. I don't know too many companies with leftover IT resources to write applications for every employee. I do think it's great to put pressure on Bill because they need to make the PC easier to use, more reliable, and more manageable.

on the job

At the end of the day, CEOs would be better off getting their IT guys to better manage their technology and application portfolio and put processes in place to manage those effectively around what they've already invested in. It's typical of the herd mentality and the hype factor of the Internet to think some new technology is going to be the magic wand. It's time management steps up and realizes that technology is a key part of the way we do business and we need to maintain this technology infrastructure with the same rigor and perseverance as we do to manage our factory, distribution, or sales infrastructures.

on the debate

Greg Blatnik Vice President Zona Research

Confusion and misinformation are the hallmarks of this debate. We've characterized it as something that pits customer need against vendor power and greed. CEOs have vendors telling them every day, "We're going to have this in six months so you need to start planning for it now." If you look at the track record over the past year, we've got a lot of unfulfilled promises, a lot of products announced that have never been delivered.

That doesn't mean there isn't merit in each of the product offerings. The NC message, at its essence, is one of easier administration, easier use, less complexity, and purportedly less cost for the organization. And what's being proposed by traditional PC companies is better administration and, in some cases, lower-cost alternatives to full-blown PCs. Those are good things.

on the job

Number one, you have to be cautious and skeptical; two, you have to ask people to put products in front of you that actually work and that you can buy today; and three, you have to have your own roadmap and have vendors prove to you that what they're offering will fit into that roadmap and won't make you take some detour that goes off a steep cliff.

on the debate

Audrey Apfel Vice President GartnerGroup

There's a big difference between network computing and the network computer, and you don't have to tie the two together. Network computing is about a new distributed software architecture. You could do it on a PC, on a phone, on a mainframe, etc. The NC demonstrates that there are viable desktop models besides the Wintel PC.

There is a place for a simpler device than a PC. The complexity of the PC environment and of distributed applications, as well as cost of ownership are astronomical problems. So the idea of a simpler device really resonates with users. People like Scott and Larry have decided to tap into that and piggy-back a whole new computing architecture on top of it.

We think there are no PC killers for the next three to five years, but some form of network computing device, or thin-client, will find its way into some pockets in most enterprises.

on the job

The first thing is to separate your network computing software architecture decisions from desktop device decisions, and think where terminal devices would be appropriate. You want to look for a set of users with a known application portfolio.

Our studies show there are real substantial cost savings for some of these new devices over a typical PC configuration. Chief executives have to be involved to make sure that the door stays open to some of the approaches that are ready and that do work and not get scared off by all the hype.

on the debate

Eileen O'Brien Director IDC

We see a desktop market with some folks needing terminals, some folks needing PCs, some folks needing NCs, and some perhaps needing this new Windows-based terminal. In the next five years, I think the PC will be by far the desktop of choice. Going out 10 years, it's a question mark. But the bottom line is that NCs are really good news for folks making buying decisions because they have choices.

Yes, Ellison and McNealy were way out there in their projections that the NC was going to immediately kill off the PC, but they got people's attention to the cost factor of the PC. So PC manufacturers have had to step up to the bar and lower cost as much as they could and put in more management features.

on the job

Companies need to do pilot projects and evaluate their users. First you have to weed out the mobile versus the stationary users and then make decisions based on the applications these folks need to get to. Do pilots. It's a new market, so somebody like Sun, who's looking to move their Javastations, will give you a few and allow you to run them, test them out. There's an awful lot of marketing BS - and there's nothing like touching and feeling.