The Real Road Ahead Of all the industries it has nailed, the Internet has most changed infotech. Over the next few years, Microsoft, Intel, Oracle, Sun, Cisco, and Sony must steer through a massive transition. Only a couple are truly prepared. The others may not be the tech leaders of the next decade.
By Brent Schlender

(FORTUNE Magazine) – The Internet changes everything. We've heard that phrase so often in the past couple of years that it has ceased to have any shock value. Of course the Internet changes everything. Why else are Web- and Wall Street-savvy IPO zillionaires propagating like rabbits? Why else do money-losing dot.coms brokering leftover airline tickets command bigger market caps than the airlines themselves? It's as if we've all been herded into a house of mirrors, where we see only the warped and upside-down reflections of both bricks-and-mortar old-economy behemoths and smug and scrawny millennium e-startups.

The reality is that the Internet, for all its buzz, has wrought flashy but mainly superficial change in most industries so far. It has transformed back-office procedures for how goods are inventoried, bought, and sold; it has served as an intriguing new medium for marketing and selling to consumers; and it has even streamlined investing so much that millions of average Americans, once sheep led by stockbrokers, are now trading stocks daily--no, hourly. Yes, buying an obscure book or auctioning off Grandma's coal-oil lamp has never been easier, and e-mail has vastly improved internal communications at most companies. And over time these e-changes will grow to seem more fundamental. But those who proclaim that the Net is a revolution that has already ripped through American business mistake a clear view for a short distance. What we've seen up to now is just an overture: It will take years for this saga to play itself out.

One big industry, however, is already feeling the full force of how profoundly disruptive the Internet can be. It's the one that brought us the Internet in the first place. I'm talking about what we used to mean when we spoke of Silicon Valley. Remember? Chipmakers, software publishers, networking gearheads, computer-box makers, consumer electronics gadgeteers, and other employers of genuine geeks. It's a collection of substantial businesses whose very real products soak up about 50% of the capital spending of American private enterprise. The Internet has rattled the geopolitics of high tech much as the end of the Cold War reshaped the real world. It has created a radically new pecking order. IBM, Oracle, and even Microsoft and Intel (the famed Wintel duopoly) are taking long, hard looks in the mirror to figure out what they really are now.

Think about it. No longer are design and production advances in the semiconductor business driven mainly by the needs of the personal computer business and the upgrade treadmill of PC software. Instead, state-of-the-art silicon is as likely to be found in a teenager's videogame machine or in a telco's switching substations. No longer is software a shrink-wrapped packaged-goods game; it's an ephemeral product usually delivered online, and increasingly sold by subscription or as a service.

The PC just isn't the focal point anymore. It is but one of a growing constellation of digital devices ranging from mainframe-like servers to networking switches and routers to TV set-top boxes and game machines to hand-held personal organizers, smart cell phones, and interactive pagers that are the vehicles for innovation going forward--not to mention the vessels for most of the profits. In short, the extended IT industry is more than ever before beginning to resemble--drum roll, please--a network.

"We're in the midst of a sea change," says John Chambers, CEO of Cisco Systems, Silicon Valley's hottest maker of networking gear. "The Internet has transformed the IT industries from a collection of oligopolies into a more diverse ecosystem in which the important thing isn't so much what technologies you own as how well you can work with the other players." Intel Chairman Andy Grove puts it more bluntly: "Business is a lot more complex for all of us now because the Internet has made the computing environment so much broader. It's not clear anymore who you need to be at loggerheads with."

The changes thrust on the IT industries by the Internet don't really represent a historical dislocation. Instead, they are the acceleration of a four-decade trend of steady proliferation in the number of key movers and shakers in IT. After all, in the beginning--or at least in the 1960s--there was effectively only IBM, which made the silicon, the software, and the Frigidaire-sized CPUs that constituted the nerve centers of corporate America.

In the 1970s, IBM was joined by a gaggle of minicomputer makers led by Digital Equipment Corp., which engineered a Big Blue-like soup-to-nuts array of IT products on a smaller scale to meet the needs of corporate divisions and midsized companies. In the 1980s yet another generation emerged, thanks to the grassroots insurgency that became the personal computer business. While Microsoft and Intel shrewdly figured out how to define the technological parameters of the new industry, their "clonable" PC architecture empowered companies like Compaq, Dell, and Hewlett-Packard that moved into the ranks of IT kingpins.

And now, in the Internet Age, the number of IT big shots is further multiplying by virtue of the all-embracing nature of ubiquitous networks. The IT industry is no longer a simple value chain with a handful of key links and choke points, but instead is a value "web" built around universally agreed-upon network standards. Because most of the standards aren't owned by any one company, there are many different potential configurations of devices and technologies to handle typical data-processing and communications chores. Beyond that, IT has begun to subsume parts of the telecommunications and media businesses too, drawing such leviathans as AT&T and Disney into the ranks of digital powerbrokers.

That's why this is such a strange time for Microsoft and Intel. They have enormous strengths, including great technology labs and the cash to acquire hot companies that develop the stuff they wish they'd discovered on their own. But the Zen-sounding slogan of their chief philosophical rival, Sun Microsystems--"The network is the computer"--now applies to even the Lords of Wintel.

If they don't watch it, the old guys run the risk of being the General MacArthurs of infotech, not dying per se, just fading away as industry powerhouses. The IT business is profoundly different, and for most companies profoundly better, than it was a mere five years ago, before the Internet opportunity presented itself. For the IT leaders profiled below, however, the Internet is as much a challenge as an opportunity: Whether they like it or not, it has changed the rules and redefined their futures.

MICROSOFT

The Network Is Indomitable. Microsoft founder and CEO Bill Gates has often mentioned that one of his goals is for his company to do what neither IBM nor DEC was able to do--namely, remain dominant while making a graceful transition from one computing era to the next. And at first blush it looks as if he's done it, because Gates & Co. has so far managed to remain the single most influential company in IT.

But if you look a little more closely, you can see that Microsoft's influence isn't as universal as it once was. Indeed, it took all the effort the company could muster to beat back the challenge posed by Netscape and its Navigator Web browser, the "killer app" that fomented the Internet revolution when it hit the market in 1995.

The "browser wars" reprised Microsoft's basic strategy of more than two decades--to identify a technological choke point for which it could design proprietary standardized software to impose conformity and compatibility onto the entire PC industry and then, of course, charge computer makers and users royalties for the privilege of such standardization. It did this first with programming languages, then with PC operating systems like DOS and Windows. Later versions of Windows also controlled how departmental local area networks of personal computers interacted. And when the Web emerged as the multimedia manifestation of the Internet, Microsoft identified the browser as the choke point it would use to control the destiny of the mother of all networks.

But this time things haven't worked out quite as before. Although Microsoft's Internet Explorer is now, by some measures, the dominant browser, its popularity has come at a price. Or perhaps more aptly, without a price. The company had to give it away to wrest market share from Netscape. So Microsoft's hottest new product in years generates virtually no direct revenue.

But even more profound and unsettling is that with the browser as interface, any computer device, regardless of its hardware or software underpinnings, can tap into Internet servers and call up the data, content, and even software applications residing on those heavy-duty computers. In other words, the Net isn't as much about strict software and hardware compatibility among devices as it is about the increasing diversity of devices that can plug in and browse.

Consequently, for Microsoft to maintain the level of influence to which it has grown accustomed in the Internet era, it must find ways to insinuate its browsing software into a wide array of devices, not just PCs, but also gadgets like videogame machines, TV set-top boxes, cable controllers, hand-held computers, and cell phones. Problem is, makers of those devices know this too, and want to develop their own browsers rather than pay tribute to Microsoft. Browsers are a relatively trivial technology that's easy to replicate, at least compared with operating systems. And unlike operating systems, which have hooks into all of a user's or a company's other software, browsers are strapless software--you can trade from one to another with little penalty.

Moreover, Microsoft also realizes that most of the real action on the Net takes place in servers. Hence, it now must also make it easier to develop Windows programs that can be "hosted" by these large computers that will remotely "serve" applications and data to these many devices. Not only does this approach challenge Microsoft's traditional business of selling packaged software for each of millions of PCs, but it also lends credence to competitors' claims that in the Internet era, proprietary operating systems are less relevant, given that most computing chores are handled elsewhere on a network.

Indeed, the server--which can range from a pumped-up desktop PC to the jumbo computers that used to be called mainframes--is the center of gravity for the Internet. Microsoft's current operating system software for servers, called Windows NT version 4, is fine for smaller machines that handle a few hundred clients, but it can't match the power of Unix on bigger iron that runs most heavy-duty Websites. Microsoft's next-generation version, Windows 2000, will supposedly remedy the shortcomings--but it won't ship until December at the earliest, practically a year later than the company originally planned. In the meantime, seemingly out of left field, yet another server operating-system rival has emerged in Linux, a variant of Unix with the tantalizing feature of being virtually free.

Those are some of the reasons Gates and his chief sidekick, Microsoft President Steve Ballmer, now concede that the browser wars with Netscape were just the start of what will eventually be a complete transformation of the company. Says Ballmer: "Some people say we met the Internet challenge head-on and already turned our ship around. Well, that's not really true. We turned the ship around some, but we're not at the end of a process. We're still at the beginning." (For more on Ballmer and Gates, see "The $100 Billion Friendship.")

Gates still argues forcefully for his vision of "symmetric" network computing, in which Windows permeates all aspects of the Net. To achieve that, Microsoft offers three different versions of Windows: Windows 98 for individual PC users, Windows NT (soon to be Windows 2000) for servers and power users, and Windows CE for hand-held computers, game machines, and set-top boxes.

Says he: "You'll still want a symmetric architecture between what goes on at the server and goes on at the PC or other client device. With that, a salesperson, say, could download part of an application and its data and use it anywhere. Our technology empowers the individual. It is a strong counterpoint to those who use the single-point-of-failure model, where if the server that does everything goes down, you can't do anything." If that sounds awkward, it's because even as keen a strategist as Gates has a hard time giving up the arguments that have served him so well so far. He's trying out yesterday's arguments on tomorrow, and they don't quite match up.

But neither Gates nor Ballmer kids himself that Microsoft's business will ever be the same. Says Ballmer: "How do you prepare for the fact that our business model is going to change fundamentally over the next five years? That's just the nature of what's happening with the Internet.... The good news is there really is new opportunity here."

What Microsoft does with the opportunity is the question that he and Gates have yet to answer.

INTEL

The Pentium Isn't the Network. The other half of the Wintel duopoly faces much the same challenge as Microsoft. But Intel Chairman Andy Grove seems a bit more reconciled to the way the Internet has transformed things. Says he: "The center of gravity of computation is shifting back into the cloud and away from the desktop. That slogan Sun uses--'The network is the computer'--is right. I wish I'd thought of it."

The "cloud" Grove speaks of is the swarm of servers that contain the Websites, the data, and, increasingly, the hosted applications that are the heart and soul of the Internet. The trend is not a good one for the company. For starters, far fewer servers are sold each year than PCs. Second, the fact that most Internet action takes place on servers encourages development of cheaper and cheaper client devices, many of which don't need a powerful chip like Intel's Pentium. And third, Intel's own microprocessor for high-powered servers, code-named Merced, is coming to market later than expected. As each month passes, more and more potential customers opt for Unix servers, which often run on chips from other manufacturers.

Moreover, Intel doesn't make the kinds of low-power processors that go into hand-held computers and cell phones, so it's missing those fast-growing markets too. And only lately did it announce its intention to get into the hotly competitive but fast-growing market for programmable telecommunications processor chips--the ones that companies like Cisco and Lucent use in their Internet routers and switches. It has a processor ready to compete in the market, but it is so late that it must play catch up to IBM and Texas Instruments.

Says Grove: "We have to face it. The dollar growth potential of microprocessors for desktop PCs is leveling off. It's not going away, but if it were still growing 30% a year, our enthusiasm for these new networking chips would be significantly less."

Intel is one of Silicon Valley's most active venture capitalists and invests in all types of nascent Internet plays--but most of them are an effort to extend its old core business by driving demand for faster processors. In late September, Intel announced it would begin running data centers, which provide human resources, accounting, inventory planning and control, and other outsourced basic data-processing services over the Net to small businesses. That's not quite the same as running some of the world's most advanced semiconductor plants though.

Grove is convinced that Intel will benefit by learning more about how the Internet will transform conventional businesses and what the technology demands will be. Still, one has to wonder whether the forays will turn into businesses that Intel can dominate as skillfully and profitably as it has the microprocessor game.

The last time Intel had to change horses was in the early '80s, when, facing relentless pressure from Japanese manufacturers, it pulled out of making memory chips and concentrated on microprocessors. It was a gutsy gamble, but it paid off because it gave Intel a 15-year near monopoly over the key component of a phenomenally popular product. None of the new businesses Intel is now dabbling in holds that kind of promise.

Concludes Grove: "All of a sudden it's a multifront war. And the stakes are bigger than ever because the entire world's commerce and warehouses and distribution centers and communications are all going to run on the Internet infrastructure. We'll try anything that makes business sense to be a part of that." In other words, Intel's days as a one-trick pony are behind it. They'd better be.

SUN

The Server Is the Computer. The story of one of the biggest beneficiaries of the Internet phenomenon is also the story of its weakness. In the 1980s, Sun Microsystems carved out a lucrative business by building high-powered engineering workstations. These supercharged desktop computers could be linked up in networks so that teams of chip or car or software designers could concoct new products using computer-aided-design (CAD) software. For a time Sun workstations were the Ferraris of desktop computers.

By the mid-1990s, however, Wintel PCs, which were much cheaper than workstations, got hardy enough to handle many of the design tasks and began to eat into Sun's market. Fortunately for Sun, the company also had a side business of machines that were then called "file servers," which acted as online archives for the networked engineering workstations.

Every workstation and server that Sun has ever built used Internet networking protocols to share data and programs. Consequently, Sun's file servers were the natural choice for companies that began to roll out Internet networks in the latter 1990s.

So even though Sun was forced to retreat from its original line of business, its retreat put it at the center of the action in the Internet revolution. And so far, it has kept the Wintel juggernaut at bay, advancing its own Unix-based server architecture while Microsoft and Intel struggle to bring their belated next-generation server technologies out of the labs and into the marketplace.

"We're the only ones who haven't tried to be anything different than what we've always been," says Sun CEO Scott McNealy. "But it's way better to be ahead of your time and let the market come to you than to have to scramble to catch up."

Sun has also made life more complicated for Microsoft by promoting a couple of other networking technologies aimed at broadening the reach of the Internet. Sun's Java programming language and Jini networking technologies will make it easier to plug all sorts of devices into the Internet. And Java and Jini also have been great marketing gimmicks. Not only have the J-technologies won Sun lots of attention amid the cacophony of Internet hype, but they have also diluted the importance of the PC by demonstrating how a diversity of special-purpose devices--printers, cameras, stereo gear, security systems, and the like--becomes even more useful when linked to the Internet.

Bill Joy, a Sun co-founder and its chief scientist, contends that the special-purpose devices and other non-PCs are what will make the Internet truly useful to people. "We're not trying to crush the PC--that's a negative goal," says Joy. "We're just trying to find a better approach to computing. We need simpler, friendlier computers, not just an ongoing cycle of more and more megahertz."

Noble words. But Sun can afford a little noblesse oblige. That's because the more devices there are out there, the more servers will be required to feed them data, software, and who knows what else. And servers are what Sun sells better than anyone else.

CISCO SYSTEMS

The Network Is Our Oyster. Nobody really owns the technology that holds the Internet together. It's a series of commonly agreed-upon, arcane standards--called IP protocols--that dictate how chunks of information are shipped around an ad hoc network. While the protocols aren't proprietary, some companies have been working with them longer than others, and the granddaddy of them all is Cisco Systems, the leading maker of Internet routers and data switches, the machines that ensure that e-mail and data get from point A to point B on the Internet or a corporate network.

"IP protocols are the hub around which this whole industry turns now, not any parti-cular operating system or platform," says John Chambers, Cisco's CEO. "I wish I could take credit for it, but we were at the right spot at the right time, because Cisco's founders built our technology on those open Internet standards. And now here we are at the center of the hub."

The first of the really hot Internet securities, Cisco has made good use of its rich stock as currency by letting Silicon Valley startups do its R&D and then acquiring them just as their technologies are ready to go bigtime. Indeed, Cisco has demonstrated that in the fast-moving Internet era, the best way to build a company up quickly is by acquisition.

Now the Internet infrastructure is growing so quickly that even today's beefed-up Cisco, with annual revenues of $12.1 billion, has to team up with others to keep up with all the opportunities. Cisco counts among its strategic partners Microsoft, Sun, IBM, and Intel, even though those companies are competitors in other arenas. That's why Chambers likens the Internet to an ecosystem. Says he: "In an ecosystem there's a need to partner closely and strategically instead of just loosely. We don't have a choice. It'll be a challenge because partnerships are more difficult to manage than acquisitions. But this new emphasis on partnerships makes this industry different now, and more like a network than ever before." These, too, are noble words. The noble words of the master of the network.

ORACLE

The Basement of the Network. For still other IT companies, the Internet era offers a second chance. In the mid-1990s, Oracle, the leading maker of database software, struggled to build a business providing corporations with back-office software applications to manage finance, human resources, enterprise resource planning (ERP), and the like. This heavy-duty software resides far from users' desks, hidden away on banks of machines in corporate server rooms, where it crunches data and streamlines decision-making for Fortune 500 companies. SAP and PeopleSoft sold the stuff and grew by leaps and bounds, but Oracle couldn't really get off the ground.

Then came the Internet, and it dawned on Oracle founder and CEO Larry Ellison that he could deliver the same kinds of applications via some server-based software and an Internet browser, and that his system could cost much less than his competitors' conventional ERP and HR software. What's more, he reasoned, the enterprise software applications could even be sold as a service rather than as turnkey systems. Thus was born Business Online, Oracle's stab at building a business around hosting enterprise-management software. Says Ellison, who can't resist a chance to take a dig at Microsoft: "It's the Internet everywhere, stupid. Not Windows everywhere."

Business Online launched just a couple of months ago but has already signed up customers ranging from small businesses to a division of General Electric, says Ellison. Oracle also is helping Hewlett-Packard start up a similar application hosting service. As you can tell by Intel's announcement of yet another such service, the idea is catching on fast.

Like Sun, Oracle is retreating to a traditional strength. All these hosted applications manipulate the kind of data customers keep in Oracle's bread-and-butter databases. When Ellison saw how successful some of his newer database customers--Amazon.com and eBay, among others--had become by exploiting his software to enable Internet e-commerce, he became convinced that the Internet infrastructure could also handle outsourced, mission-critical corporate software services. Says Ellison: "In the Internet era, big is better, and we do big best."

Oracle's stock is up 143% in the past year--which sounds good until you realize that eBay's is up 1,050%, and Amazon's, 297%. Oracle may be on the right track, but it has to make sure it gets a lasting kick out of the party it's helped create.

SONY

The Network is the Living Room. When the world computer industry gathers in Las Vegas this November for its annual fall Comdex confab, Sony CEO Nobuyuki Idei will join the likes of Bill Gates and Scott McNealy as one of the event's keynote speakers, marking Sony's initiation into the fraternity of computer industry movers and shakers. Indeed, because interactive audio and video are now such important components of the World Wide Web, Sony is becoming a big player in just about every aspect of the network world.

All you have to do is scan recent business headlines to see just how serious Sony is. In mid-September, Sony announced a $1 billion deal with Cablevision to help roll out an interactive broadband cable TV system in New York City. A couple of days later the company said it would invest in TiVo, a Silicon Valley maker of a new kind of set-top box called a Digital Video Recorder, which records TV programming on a computer hard disk, for replay at a user's convenience. And then two days later Sony unveiled its first Web-oriented Walkman, which, via a PC, downloads digital music from online music stores. All that in just one week.

In fact, Idei has mobilized the entire company to make its products plug and play on the Net. For example, as of this fall, all Sony PCs, digital camcorders, and audio and video components will feature special digital sockets that let them be linked by an Internet-compatible technology called I-Link into networks that will make controlling them via computers or other Internet devices easy.

That means you could use your PC at the office to control an I-Linked home-security videocamera on your front porch to see whether a UPS shipment shows up. The same I-Link network would make it possible to use a Net connection to play music on a desktop PC in your den (or your office, for that matter) from a CD in your 200-disk carousel in the living room. If the kids were visiting Grandpa, they could log on through his TV to show him digital photos and video clips of their first trip to Europe--images stored on the family PC back home.

Then there's the PlayStation, Sony's videogame player console, which has been the hottest consumer electronics gadget of the 1990s. It is such a big seller--Sony moved 20 million of them this year alone--that it is having a profound effect on the capabilities of the semiconductor industry, giving chipmakers the incentive to go for broke to build the hottest graphics processors ever.

Indeed, Idei is so convinced that the Internet is the key to future growth for Sony that he uses the network as the organizing metaphor for the company. Now Sony is divided into five main divisions, all but one of which have the word "network" in their name. Says Idei: "We have finished the transition from analog to digital in all of our product line, and now at the millennium it is time to take consumer electronics onto the network. It's just as big a job as going digital, and it will take some time, but it is inevitable."

Ultimately, if Idei's plans work out, a large proportion of the devices hooked up to the Net won't be personal computers per se, but personal gadgets. And the name Sony, which is synonymous with the world's best TVs, portable music players, and videogames, could then be synonymous with online fun too.

There are other IT companies for which the Internet has changed everything. Looking like the walking dead three years ago, Novell, a maker of networking software, has come roaring back by exploiting demand for software that keeps track of who's who on networks. And there's IBM, which has reinvented itself as the e-services company. IBM employees don't dare utter the M-word (mainframe) anymore.

As might be expected, IT infrastructure companies are among the earliest and most serious adopters of e-business techniques. Dell Computer brags that it will soon sell more than half its computers via the Internet. Cisco requires all its suppliers and resellers to transact all business with the company over the Web. And just about every software maker now gives customers the option of downloading bits rather than ordering them in shrink-wrapped packages, lowering already microscopic production costs to nearly zero.

But the most interesting fallout from the Internet phenomenon is that it has made the IT industry seem open-ended again. There's a new sense of competitiveness, and more ways for smart companies to succeed, because this really is one of those sea changes in the industry. Only a few years ago it seemed inevitable that Bill Gates and Andy Grove and their successors at Microsoft and Intel would rule the IT world by fiat for the next couple of decades. But as strong and shrewd as Microsoft and Intel are, the Internet is beginning to show that they, like IBM and DEC before them, are mortals like everyone else. In fact, there's no guarantee that if, in ten years, Fortune decides to round up the key technology leaders for another such article, Microsoft and Intel will be among the select group. They both have peerless technology teams and the cash to buy their way into the Next Big Thing (if they don't create it themselves), but that's just how uncertain the future of technology has become.

Says Oracle's Ellison: "Just as surely as IBM is no longer the center of the universe, they never lost their monopoly on mainframes. It's just that mainframes ceased to be at the center of the universe. The same thing's happening to Intel and Microsoft and the PC now. And they're finding that it's definitely more complicated when customers have a choice and you have to compete. In fact, it's a bitch. You need to build multiple products, and hedge your bets, and sometimes someone beats you. These things happen. It can be very humiliating."

Yep, the Internet has changed everything. In the IT world, at least.