Anadigics makes the case for GaAs

Electronic News, Feb 9, 1998 by Gale Morrison

When Anadigics' Ron Rosenzweig, chairman/CEO, and Charles Huang,executive VP and COO, met with Electronic News editors, the company was not under attack by Wall Street as it was last week. After reporting earnings that showed revenue and unit increases but included a dire warning--the current quarter might see a net loss due to "a substantial reduction in orders"--the company's stock took a hit, falling 58 percent in one day.

As a producer of gallium arsenide semiconductors, Anadigics' gallium arsenide devices face new competition from silicon germanium, which threatens to displace GaAs in many communications applications and which has the deep pockets and drive of IBM behind it (see story above). Still, the company has its sights on markets that require tens or hundreds of millions of chips a year, as opposed to the thousands demanded by the military business with which it started.

What became clear in talking with Messrs. Rosenzweig and Huang is Anadigics' number one priority is a profitable business with a strong customer base, and missionary work on the benefits of gallium arsenide comes second.

"We've developed a gold-plated customer base within the communications-markets area and if we think we can serve them with silicon circuits, or silicon germanium, or other types of gallium arsenide, then our intention is to do that in the most cost-effective way we can, and try to provide more value to the customers," Mr. Rosenzweig told EN.

As attendees of the Wireless Symposium opening today will bear out and Mr. Rosenzweig said, in many areas of wireless communications "(t)here are no Intels." IBM and Texas Instruments have units that run in the same pack as RF Micro Devices and Anadigics. Only time and technology will tell.

EN: How has Anadigics changed?

Mr. Rosenzweig: The company started in 1985 with about six or eight people at a time when there was tremendous interest (in) and excitement about gallium arsenide. Most of that interest was driven by defense applications, some high-end communications applications and supercomputers.

In the period from '85 to '89, we ... basically established our technology, and pursued some more traditional applications. But the real Anadigics story started around 1989, when we concluded that there were mass-market opportunities, for gallium arsenide radio-frequency and communication-centered ICs. And the key paradigm shift was the . . . conclusion and commitment we made that you could make these products for $5 apiece and make money. And you could build a very large business by selling millions of them, not thousands of them. The old paradigm saw thousands of chips each selling for $100; our paradigm was millions at $5 each.

The first successful product was a chip that we developed for the direct-broadcast satellite-TV industry as it was starting in Europe, driven by Rupert Murdoch. The European marketplace had four local channels and nothing else. Because there was no cable, or very little cable, DBS became an instant success and we wound up being a supplier to about 50 to 60 percent of the marketplace. During the '89, '90, '91 time period, we proved that we could sell ICs at $5 in million-piece quantities, and actually changed the way the market dynamics worked.

And the model that we pursued there is the model that we've used for all of our markets: we essentially displace discrete components. We were a discrete-component killer.

We displace gallium arsenide discrete devices, silicon bipolar discrete devices, RF and microwave diodes, as well as a number of other passive components that were needed to make these types of circuits. The reason that this was a discrete-component business, and not . . . an IC was simply that silicon-integrated circuits then and now could not meet the performance requirements of these 1,000-, 2,000-, 10,000MHz types of applications, factoring in all of the other efficiencies and sensitivities and dynamic ranges that were required by these types of products. (It is to this market area the new silicon germanium is targeted.)

After the first success with DBS, we developed products for fiber-optic communications and we continue to expand that product line.

The next big mass-market product was for cable TV. We teamed with General Instruments and developed chips that go into the . . . tuner section of . . . a set-top box. It is the chip that's used to receive the front-end signal, the wideband 500MHZ--or 800MHz-wide signal, and then find channel 2, channel 10, channel 8, or whatever.

In the '92-'93 time period, we saw tremendous opportunities in the rapidly-expanding cellular telephone market which, at that time, was dominated by discrete-component solutions or, in some cases, in a hybrid-circuit form. We teamed with Ericsson in that work, just when they were beginning what's turned out to be tremendous growth rate in the cellular market.

In the cellular telephone market, we addressed the receiver side, for that was the first communications market that required a combination transmitter and receiver, in effect, a transceiver. Then we directed our technology towards the transmission side and developed ICs that take a 1 milliwatt signal and boost it to 1 watt for a cellular telephone handset. And we did that first for . . . the analog bands in the United States, the "hams" band; also the European analog; and then started to develop products for the new digital-communications bands. "And that market, expanded now by the PCS market (that) was created by the auctions in the United States, and the opening up of the 1,800MHz to 2,000MHz frequency bands, has just been the most explosive market that any of us in the RF/microwave world have ever seen. In 1997, 100 million telephone handsets were sold. And that market is growing at about 40 to 50 percent a year. We estimate that our share of that market is between 15 and 20 percent of the total . . . available world market, up from about 7 percent the previous year, and 5 percent the year before that.