Wintel: End of the road?

The brave new world of digital convergence is coming--the question is who will make money in it.

One of the most obvious differences from the prevailing digital world of today centered on personal computers will likely be the absence of Intel or Microsoft from a leading role--at least initially. This should engender business and pricing strategies far different from anything seen to date in the PC industry.

"[Windows-Intel] is not likely going to be as dominant in the appliance paradigm," said Steven Milunovich, a vice president in Merrill Lynch's technology research division. "Processor rivals now have a chance."

Instead, the old hardware guard may find itself chasing low-cost, low-powered designs based on chips from companies such as Advanced Risc Machines, Mips, Hitachi, and electronic goods specialist Sony, as well as a host of unknown chip vendors.

On the software side, Sun Microsystems and Network Computer also stand to do well. Microsoft could make its mark in this space as well.

The wild cards will be a host of start-up companies targeting convergence, such as TiVo.

For years, PC powers have tried to expand their presence in the home. The problem is that price points have been typically high, as with PCs. Expensive PC-TV theater systems from Gateway and Compaq floundered in the market. Compaq terminated its PC-TV line, while Gateway has recently dropped the price on its existing line to make it more attractive.

In their stead, companies like emachines, backed by South Korean PC giant Trigem, have emerged to offer $499 PC-like devices that connect to a TV and offer and Internet access--and don't necessarily offer Windows or Intel processors.

In this world, Intel and Microsoft, will run up against a new market model where diversity of functions and price points work against quick standardization, leading instead to a pack of leaders, Milunovich believes.

Of the two, Microsoft will likely have an easier time fitting into the new markets. Microsoft already has established partnerships for marketing Windows CE as well as a beachhead with WebTV. The desire for PC compatibility will also help CE's appeal.

By contrast, Intel has only begun its device push. "Intel has almost no share in the appliance market. Intel is not happy about that," said Milunovich.

Intel's push into this market, based on the StrongArm processor, will give it an entry, but it comes with a catch. Because the design gets licensed from Arm, Intel does not keep all the profits from the chip.

Even the winners in this market, however, will not have an easy time with the price points.

Initially, the winners and losers in this market will largely be determined by ISPs and other service providers because

Intel's set-top stab How the chip giant breaks down the market
processor	functions	speed	chip price*
StrongArm	handhelds, basic set-tops with cable connection, and limited Web access	190 MHz	$33
Celeron	More advanced set-tops with full Web access, games, telephony, DVD	266 to 333 MHz	$86-$192
*Price is official wholesale price. Actual price for computer vendors, and retail price, will vary. Source: Intel

they will act as the consumers, said Ashok Kumar, computing analyst with Piper Jaffray. These companies will buy these devices en masse and then recover the cost from customers.

This buying scenario, however, puts the vendors at a disadvantage. "The number of service vendors are few. They will exert a significant amount of pressure on the hardware vendors. The silicon opportunity is maybe 50 or 60 bucks," he said. "Gross margins will be less than 20 percent."

Nonetheless, chip vendors will try. "Anything other than [memory chips] is good for them," Kumar said.

Also, Japanese companies may stake a large claim in this market since they are already a significant force in the consumer electronics world.

"The Japanese game [machine] manufacturers have brand recognition, they are very good at mass marketing, and they have low-cost manufacturing and customer service," said Jim Turley, embedded processor analyst at MicroDesign Resources.

The changes the convergence market will engender, especially for PC warhorses such as Intel, Microsoft, and Compaq, come largely as a result of the cost realities of consumer electronics.

"Smart" set-top boxes that include a TV cable connection, a game-playing platform, and some form of Internet access will likely come in at a price point of $299 or less, many say. The chips that power into them will sell for $30 to $60, well under the average price of about $200 for today's PC processor. Features such as DVD drives or an ability to perform computing tasks will raise prices, but likely not get them above $1,000.

With such a tight envelope, success will lie in efficiency, strategic control over intellectual property, and volume sales.

On the system side, currently Japanese vendors look to be in the lead because their game-playing devices are almost at the stage where they can serve as intelligent set-top boxes.

The Sega Dreamcast will take the process one step further when it comes out in 1999 by incorporating a Web browser and a modem. It also will contain Windows CE, which will allow for PC compatibility. At that point, the only crucial feature missing will be cable connectivity.

Although they can cost-effectively manufacture set-top boxes, American vendors such as Scientific-Atlanta are at a disadvantage because they do not have the game platform.  

Go to: How chips stack up