The best-selling item in consumer electronics now is the wireless phone. It seems ubiquitous.

True ubiquity, however, is Kleenex. Not wireless phones, which are used by only about 30 percent of the U.S. population.

But every expert believes wireless is on the verge of omnipresence. Over the next decade, most households are likely to have a wireless phone, or a wireless personal digital assistant (like the Palm handheld computer) or a wireless something, or maybe two or three of them, as the technology moves ahead toward transmission of faxes, stock trades and e-mail requests for money from your kid at college, which you will then be able to fork over digitally with the press of a button.

Like tissues, we use these devices and then throw them away.

Dispose and replace. Dispose and replace. That's nirvana for investors.

But the market knows everything I'm telling you, and the prices of the wireless stocks reflect that--when it comes to phones meant primarily for talking.

What the market can't gauge is the speed with which consumers will embrace wireless phones and other handheld devices equipped for data, the "third generation" of wireless. It will happen. (The dramatic run-up in the stock of wireless phone maker Qualcomm is in part due to its patent rights to the transmission standard, called CDMA, considered most useful for data.)

But those Internet-ready phones you see advertised are primitive. Estimates of how long it will take this technology to evolve range from two years to 10.

In uncertainty there lurks opportunity. The third-generation devices--3G, as it's known--will have to be faster and far more efficient at processing radio signals with low power consumption. Who wants to wait forever for a download and then have the batteries run out in the middle?

The companies working to advance the technology of efficiency are already very busy and will become more so.

I'm talking about component suppliers. Nokia may make phones, but it buys the chips and processors and filters that go on the phone's electronics modules from companies you may have never heard of.

These components are "the enablers" of the wireless evolution, says Jane Zweig, executive vice president of the Wheaton-based telecommunications consultancy Herschel Schosteck Associates. They "are very important to any terminals that will deliver any of this stuff, probably crucial to the value chain," she says. "You can have the cutest terminals, but if the chip's not good, terminals won't deliver."

Let's back up a bit. Thanks to controversies about whether to allow cellular antennas in Rock Creek Park, most of us know by now how cellular phones work. When you dial your phone, it sends a low-energy signal to an antenna, which relays it to a land line and on to the number you dialed.

To function, a digital wireless phone needs, among other things, microprocessors and chipsets to receive signals, decode and filter them, and send them out again. To be practical, it also needs a special chip, called a radio frequency amplifier, to transform pathetic little outgoing signals into more powerful signals that will reach back to the antennas rather than just dribble out of your phone.

We use a great deal of power when we talk. The batteries run down fast--at the crucial moments of our conversations. Data transmission requires much greater speed, running on the same or lower battery power, with smaller components to allow room for all the extra components required for data.

Well-known companies are hard at work on this challenge, Motorola, Lucent, Cisco and Texas Instruments among them.

But busy also--well beyond their capacities, at the moment--are lesser-knowns such as RF Micro Devices of Greensboro, N.C. (revenue $153 million for fiscal '99); Anadigics Inc. of Warren, N.J. (revenue $150 million for the first three quarters of fiscal '99); TriQuint Semiconductor Inc. of Hillsboro, Ore. (revenue of $112 million for the first three quarters of fiscal '99); and Alpha Industries Inc. of Woburn, Mass. (revenue of $126.3 million for fiscal '99). This list is, by design, illustrative rather than exhaustive.

These companies have in common great year-on-year increases in sales as wireless has taken off, dramatic run-ups in stock price and high price-to-earnings ratios. But there's a reason they're so valuable: They develop and manufacture gallium arsenide chips, originally used by the military, which allow for a much faster movement of electric current than silicon--roughly six times as fast--using the same or less battery power. Bell Labs research is speeding them up further.

"The difficulty is for us to keep up" with demand, says Greg Thompson, director of sales at RF Micro Devices. "Growth is astronomical."

I don't mean you should rush out and buy these stocks. Look deeper. There's a drawback to semiconductor stocks generally. "I think they're very risky investments," says Robert Gensler, vice president and telecom analyst at T. Rowe Price, the Baltimore-based mutual fund company. "Everybody has component shortages right now. . . . But at some point the capacity will catch up with demand." Then, he says, "the end guy, Nokia, or Motorola, will still be great but the components guy hits the wall."

Wireless data is a new challenge, however, and no one can predict the point at which capacity will catch up with demand.

A Cambridge, England-based company, ARM Ltd., makes the design for the microprocessor cores that go into 70 percent of the wireless phones in the world today. It, too, is known for developing ever-more-efficient processors of the sort necessary for the next decade.

The list of ARM-powered products that are publicly known (some companies would just as soon keep it to themselves) would take up the rest of the page. A few examples may be familiar: phones made by Sony, Toshiba, Sanyo, Hitachi, Ericsson, Nokia and Kyocera; cameras made by Toshiba and Sharp; games made by Sega, Samsung and Sony.

ARM's stock, sold in the United States through an American depository receipt (ARMHY), has had a run-up (843 percent in the past 12 months) as staggering as its P/E ratio, nearing 700.

To some extent it's a momentum stock, the sort that people buy largely because they believe others will do so, en masse.

But ARM is also a money machine. The company collects money for each license to a manufacturer. On top of that, it collects money for each microprocessor sold using the license. It has only 400 employees, but a market capitalization of about $9 billion and revenue in the first three quarters of 1999 of about $80 million.

"Yes," ARM is "extremely expensive," says Drew Peck of S.G. Cowen. "But so are most names that are indirectly tied to the communications sector." ARM "is gaining market share at a dramatic rate," he says. "They have a unique technology particularly well suited for portable consumer products. . . . The fundamentals are profoundly strong."

Let me say emphatically these are not personal "buy recommendations." My only recommendation for forward-looking investors who are unafraid of the volatility of technology stocks is that you research wireless carefully, and soon, rather than running out and buying the companies with the best advertising. The ads are for phones and service, not stock.

Go to a Web site called How Stuff Works (howstuffworks.com), and you'll find a clear explanation of how wireless works. Go to the Web sites of Qualcomm, Ericcson, Motorola and Lucent for more information. Look at the news releases to see who these companies are partnering with. Go to a search engine called Google.com--as I did--and search for "wireless and chips," "wireless and microprocessors," and "wireless and components." There are all sorts of industry articles and academic studies, at varying levels of intelligibility.

The companies' Web sites will lead you to their Securities and Exchange Commission filings, which provide relatively blunt explanations of risk, such as the extent to which their revenue may be dependent on just one or two companies or one or two products.

If all this is too much for you, go to another corporate Web site and check out another product we replace, at the rate of 200 billion a year. The shares have tripled over the past decade and the P/E ratio is only 22.

The product is Kleenex. I can recommend that.

Fred Barbash (barbashf@ washpost.com) is The Post's business editor. Of the stocks mentioned in this article, he is a longtime owner of Lucent and Cisco Systems.

The Next Generation

Some companies that develop and make gallium arsenide chips -- which hold promise for wireless phone data services -- have seen great growth in share prices over the past year.

Consensus estimates of earnings per share; 12-month share price growth

(Current fiscal year; Next fiscal year)

Anadigics Inc. (49 cents; 86 cents); 270.4%

RF Micro Devices (59 cents; 78 cents); 1,039%

Triquint ($1.14; $1.27); 529%

Alpha Industries ($1.09; $1.44); 335.9%

Specialized semiconductor industry 107.9%

SOURCE: Zacks