Q&A With David Aldrich, President And CEO Of Skyworks Solutions, Inc.

By Jim Pomager
Editor in Chief

The advent of high-speed data and multiple modes of operation has forced wireless handsets to morph into incredibly complicated, extremely sophisticated devices. Carriers are asking handset OEMs to perform what is essentially a minor miracle -- create a phone that can handle data and voice simultaneously in an environment where signal strength, interference, and temperature are almost constantly in flux. And, of course, carriers want it small and they want it cheap.

One way OEMs are solving this problem is by leaning more heavily on their own supplier base. Semiconductor companies, for example, are responding to the challenge by integrating more components into single-chip and multichip modules, which ultimately simplifies handset designs and reduces costs for the OEM. One chipmaker at the front of this revolution is Skyworks Solutions. Founded less than four years ago, when Alpha Industries merged with Conexant Systems' wireless business, Skyworks now supplies radios to four of the top five handset OEMs (LG, Samsung, Motorola, and Nokia). And the company is parlaying this cell phone know-how into a line of linear products aimed at a wide variety of vertical markets.

I recently met with Skyworks' President and CEO David Aldrich in New York City, after he addressed investors at the 8th Annual Needham & Company Growth Conference. He told me about his 20-year history in the RF/microwave industry, about the formation and evolution of Skyworks, and about his view of the wireless industry's future. He also shared his thoughts on some emerging wireless technologies (including WiMAX) and even weighed in on the gallium arsenide (GaAs) versus complementary metal oxide semiconductor (CMOS) debate.

Click here to download the article in PDF format.

RF Globalnet (RFG): How did you get your start in the RF/microwave industry, and what experiences brought you to the helm at Skyworks?

David Aldrich (DA): I started my career in the early 1980s at Space Communications, a partnership between Fairchild Industries and Contel that provided ground support, software, and system integration for NASA TDRSS (Tracking and Data Relay Satellite System) satellites. It was fun, but short-lived. After all the satellites were launched and the ground support was in place, the company essentially went out of business.

I took my experience around satellite communications to Adams-Russell Electronics, which built a variety of components -- mixers, amplifiers, etc. -- for military applications. That company was bought by M/A-COM in the early 1990s.

For several years, I ran a business development group at M/A-COM, and then ended up running their GaAs power amplifier business. Those were the very early days of applying GaAs into high-volume commercial applications, such as low-noise amplifiers, GPS components, base station components, and switches for handsets.

In the mid-1990s, just before the big boom, I left M/A-COM to join Alpha Industries, a defense contractor that built very specialized GaAs semiconductors for satellites, missile systems, and radar systems. The U.S. military budget was not particularly strong at the time, so we made a strategic decision to drive Alpha's technology into commercial applications, namely cellular telephones.

The transformation was a pretty dramatic one. It impacted our company culture, our systems, the way we designed products, and the way we supported our customers. We discovered that by trying to serve the military and commercial markets simultaneously, we ended up serving both poorly. So, we ultimately converted Alpha from a company that was almost exclusively a military supplier in the early 1990s, to one that was almost 100% commercial by the end of the 1990s. It was a radical transformation that most companies don't make successfully.

Then, in 2002, we merged Alpha Industries with Conexant Systems' wireless business, and Skyworks was born.

RFG: What are some of the key milestones in Skyworks' history?

DA: One key milestone is the creation and continued evolution of the Helios™ radio solution. Skyworks has always combined RF, analog, mixed-signal, software, and system integration expertise under one roof. Now, we have pooled all that experience in one essentially bulletproof, easy-to-use radio system that can be used with virtually anybody's baseband -- all for about $6 per phone. LG and Samsung are currently using the Helios EDGE system. Motorola recently announced that it has adopted Helios DigRF (see related story). And we're using the same integrated radio approach for CDMA with two top-tier cellular OEMs.

In addition, about a year and a half ago we made the decision to direct more of our design, application, and systems resources into our linear products business. We took the analog functionality that we worked so hard to perfect in the cellular handset, and we applied it to derivative market applications. These applications typically have much longer product life cycles, much higher average selling prices (ASPs), and much higher gross margins. We have invested a great deal in linear products business over the past couple of years, and now we're beginning to see results with a product portfolio that will make Skyworks less volatile and more profitable.

RFG: Was that the driving force behind launching a linear products line -- to offset the volatility of the handset market?

DA: It was by no means a defensive strategy, but a way to leverage our existing capabilities. Open up a cellular handset and look at the radio system. That radio, which is sold for a few dollars by the hundreds of millions into cellular handsets, is as complicated and integrated as anything on the planet in the analog space. So as you perfect that capability, you create an arsenal of technology and manufacturing techniques that is applicable not only to the handset, but to dozens of vertical markets around the world. And in most of those markets, there is far less competition and far less customer concentration.

RFG: What percentage of Skyworks' revenue currently comes from linear product sales, and do you expect that number to grow?

DA: Today, linear products generate about 20% of our revenue, and we expect the business to grow. Will it grow faster than mobile, so that it becomes a bigger piece of the pie? We don't know, because we also have high growth aspirations for our mobile platforms portfolio. This market is very different than the handset market. For example, with linear products, we never find one opportunity that generates $20 million in revenue, as we might with a handset opportunity. Instead, we tend to find 100 different opportunities that each generate $1 million, $2 million, or $3 million. And when one of our linear products is designed into an application, the life cycle can be several years long, compared to about 18 months for a handset.

RFG: What specific vertical markets hold the greatest interest for your linear products group?

DA: The medical field is a fascinating market for us. There are more applications for this class of technology in medical than I would have expected. For example, one of the largest medical products companies in the world is using one of our simple transmitters for low data rate output to remote diagnostics from a catheter to a personal area network. As the catheter is inserted into the body, it feeds information that is used to diagnose the patient.

Digital broadcast satellite is also an exciting area for us. We have these silicon switch and control devices, fabricated in a very inexpensive CMOS, that are replacing many components and existing designs. We are beginning to sell them now by the hundreds of thousands, and we are moving into the millions.

Other market areas of interest include wireless LAN (licensed and unlicensed), wireless data, RFID tags, and automotive, among others.

RFG: Speaking of CMOS, where do you stand on the great GaAs-versus-CMOS debate?

DA: A cellular handset has three basic elements that relate to semiconductor technologies -- the baseband, which provides the digital processing; the analog, or radio; and the front-end module (FEM), which includes the filtering, switching, and amplification. The baseband element has and will continue to be done in bulk CMOS.

On the other end of the spectrum, though, I believe that the physics of the materials and the demands for efficiency and linearity overwhelmingly favor RF semiconductor technologies, not CMOS. In my opinion, heterojunction bipolar transistor (HBT) technology is hands-down the best technology available for the FEM. And in five years it will still be the best. In order to achieve the necessary level of integration and keep the cost points down, yet still handle the complexity and provide good performance, you will likely need technologies like HBT, pseudomorphic high-electron-mobility transistor (pHEMT), and RF CMOS.

I think CMOS is great for transceivers, especially for voice-centric and single-mode products, but the FEM will continue to require a compound semiconductor technology. If you use CMOS for even very low-end, voice-only FEM applications, the performance will be poor relative to GaAs, and it is patently unacceptable for data.

RFG: Everybody seems to be talking about WiMAX these days. What do you think about WiMAX, and how is Skyworks getting involved with the technology?

DA: I think WiMAX has the potential to be extremely important. What remains to be seen is how the carriers interact with WiMAX networks. Who will make money on it? As WLAN moves from a hot spot to a wider area, do traditional service providers embrace it or fight it? The industry will sort out the answers to these questions over the next couple years. And those answers will determine whether WiMAX becomes a mainstream, very high-volume market of interest to semiconductor providers like us, or becomes a small-volume, limited application.

As for Skyworks' involvement in WiMAX, we have decided not to sell radios for wireless LAN today. We think that as a stand-alone market, wireless LAN is saturated with competition, whether you're talking about PCs or other devices that sit on your desk. So in that market, we have formed partnerships with baseband providers -- Broadcom, Conexant, and others -- under which we essentially sweep in all of that passive, filtering, switching, control, and amplification functionality into a module.

But as WiMAX becomes integrated with the handset, it starts to hit the sweet spot of what Skyworks can and will do. The very existence of high-speed data capabilities in the handset facilitates lots of applications. And we believe the integration of wireless LAN in the handset is one such application that holds promise for Skyworks. So that's where we stand with WiMAX today -- supplying front-end modules to our partners and facilitating integration in the handset.

RFG: How would you characterize Skyworks' current position in the market?

DA: In the mobile environment, we have about a 40% PA share; that's up from about 30% a year and a half ago, so we're growing rapidly there. In radios, we have done very well in GSM with our Direct Conversion Transceiver (DCR) technology, and we're continuing to grow. Plus, we're one of only a handful of companies in the world that do both GSM and CDMA radios.

If you step back and look at today's mobile environment, you see what is basically a single-function (GSM or CDMA), voice-centric world. But that's changing with the emergence of high-speed data technologies like EDGE, W-CDMA, and WEDGE. As more and more phones are required to simultaneously handle voice and data, the analog portion of those phones become far more complex. The conditions in this environment are favorable to Skyworks because, as I mentioned earlier, the company was created to combine digital, software, mixed-signal, analog, and RF capabilities under one umbrella. We have 1,000 engineers with expertise across the entire radio system. Compare that to our competition, which tends to have more limited expertise.

Our recent wins with customers like LG and Samsung required strength not only in the semiconductor products that make up the radio solution, but also in the software and systems integration needed to make the solution work. We wouldn't have been able to win that business if we didn't have software and system engineering skills, in addition to components and radios.

RFG: What do you see in the future for the wireless industry?

DA: The unit volume growth has positively surprised many people who have been watching the wireless handset industry. For the first time, we will exceed 800 million units in a year, and we're moving toward a billion units per year. As the installed base increases, product life cycles are getting shorter. The number of features is always increasing, and as new features are added -- currently driven by high-speed data offerings -- there is more incentive to turn your phone over. And number portability, which is available in some markets, means that there is no downside to switching carriers. As a result, you can continually swap out your phone for the latest, coolest technology.

I'm a relatively old guy in this space, and us old guys sometimes tend to view the market from a North American point of view. We see only the features we want -- voice, data, and maybe access to e-mail. But travel around the world and look at what 13-year-old kids are doing. They're fearless with technology, and they're adopting wireless in a huge way, using it for games and sharing pictures, video clips, and music. They will use mobile phones in a way that we, perhaps, would struggle to accept. And that's the future of this industry.