Microwaves & RFs Technical Director, Jack Browne, talks with Jerry D. Neal, RF Micro Devices executive vice president of marketing.
MRF: In December 1994, Microwaves & RF ran that well-known Alexander Graham Bell photograph on the front cover, modified to include engineers from QUALCOMM and RF Micro Devices to represent your budding relationship with that company. How did that relationship start?
Jerry D. Neal: We started RF Micro Devices in 1991. One of our first design wins was with Nippon Denso in Southern California. They had asked us to design an HBT power amplifier. We did, but it had a thermal runaway problem and they ended up not using the original design. But we went back and fixed it. Last year, as you know, we shipped our billionth power amplifier.
Along the way, we worked with our sales representative in Southern California, Cain Technology, to get us into QUALCOMM. I went to QUALCOMM and got better acquainted with them and met the people developing the first CDMA phones. They had formed that joint venture with Sony to manufacture the phones. We told them that we wanted to be on the RFQs for the RF components in those phones. And at that time, QUALCOMM was not making any of the RF components, such as the LNAs or the power amplifiers. They sent out RFQs for five separate components. We felt that if we could get just one of those, it would be a big-volume job for the company. We did all these proposals, and we got all five. And it almost sunk our company because it took all the resources that we had to develop these parts, and these CDMA components were very challenging.
At one point, we took a picture of everyone in RFMD wearing QUALCOMM tee shirts. That relationship was difficult at the beginning but that's what enabled us to be the company that we are.
MRF: At that time, there was a lot of doubt whether CDMA would take off. Did that doubt ever concern RFMD?
Neal: One of my partners, Bill Pratt, and I went to QUALCOMM and received a very in-depth technical explanation of how the CDMA system was going to work and how they theorized it would perform. After Bill, who has a marvelous technical mind, heard this explanation, he felt CDMA would probably go, so we figured the probability was good that it would take off. As a young company at that time, we were taking a risk on it. But with any small company, there are risks. And the reward for this particular risk was huge.
The relationship with QUALCOMM also got us notoriety, which we wouldn't have gotten in any other way. Your magazine's coverage also helped us as a young company. It showed us that even as a young company we had some expertise and technology that allowed us to go up against much larger companies and still succeed.
MRF: The company has undergone impressive growth since those early days. Were there any challenges in managing that growth?
Neal: When we did the first QUALCOMM chips, we had about 40 people in one facility. Now we have about 2800 people, in 16 countries. We always wanted to follow a charter of making low-cost, high-volume parts and we've stuck to that. What has made it good for us is that our employees are talented and have been a huge help. We also had a culture that helped us grow.
Starting with the three founders, we tried to give our jobs away as quickly as we could. Sometimes the founders can't turn the company loose. It's called the founder's syndrome. And they want to have their fingers in every decision. We didn't do that—we hired great people and turned them loose, and we acted as consultants in many cases. The growth for us hasn't been as difficult for us if we didn't have that culture in place.
MRF: Some of the company's growth has involved acquisitions or investments, such as RFMD's investment in Jazz Semiconductor. How has that worked out?
Neal: Several years ago, we realized that even though we had started in GaAs—we are ramping up, by the way, do process the equivalent of 8000 4-in. GaAs wafers per week—that the kind of components we would be designing would require silicon. We had a relationship with IBM, but we did not have much of a relationship with the other foundries, such as TSMC—we had only a fledgling relationship with them. We decided that to really get involved in silicon, we needed a partner with a fab where we could really develop technology. Part of what we are talking about at present is having a platform where we can integrate the PA, the filters, and the switches onto a chip so that we have a true system on a chip (SoC). To do that in straight CMOS, it will be difficult to achieve the required performance. With Jazz, our strategy was to have a group that we could develop new technology with. And we have developed several unique technologies with Jazz. We have LDMOS, an MOS technology in silicon, that we think is a platform that will allow us to integrate the transceiver components and passive devices on a single chip. But we also think that CMOS is going to be very important in the future and there will be many applications that will have a CMOS PA.
We did a second thing with Jazz. We wanted to guarantee capacity, and we thought that capacity could get tight with our current suppliers. So having a benchmark capacity guarantee that we could count on. The 10-percent ownership enables us to learn and to develop processes that were custom to us.
MRF: In one of your other relationships, with Silicon Wave, RFMD actually acquired the company. How has that worked out?
Neal: Actually, we have put Dave Lyon, the CEO of Silicon Wave, in charge of our wireless connectivity business. Last quarter, that group grew by 50 percent. We are seeing some of the highest growth in the company in the Bluetooth area. I've talked to some editors who feel there is no money to be made in Bluetooth. Well, our competitor, the number one supplier in Bluetooth, CSR, for the first six months of 2005, they shipped 162 million devices at a gross margin of 47 percent. They are making a ton of money. And we think there is a lot of opportunity there. We have developed a single chip that has the whole transceiver, the processor, and memory. We have prototypes now, and will be in production in 2006. We are actually doing a couple of combinations of Bluetooth chips. We are already shipping chips to Motorola without the processor. And one of the new variations includes near-field communications with the Bluetooth capability.
MRF: The company has done well in cellular communications, and Bluetooth is growing. But what happened with the wireless-local-area-network (WLAN) efforts?
Neal: We announced that we were discontinuing our WLAN chip-set development. We don't really have a history of giving up on anything, so that was a painful decision to exit WLAN. We didn't think there was ever going to be an opportunity for us to make money there. We are accustomed to 15-to-20-percent price erosion in our markets every year. In that market, those prices were eroding 50 percent per year, those are very difficult componentsto make and, frankly, we were behind the main suppliers. What we decided to do was to exit that but stay in the WLAN business and do the frontend module, with the PA, the switches, and that would enable us to leverage our strengths and go out to our former competitors, such as Atheros, and work with them to sell them something that they need. And that business model has been well embraced. WLAN was a painful experience for us, but we learned something from it, and now we have niche that we can work in.
MRF: Has WLAN has been a difficult market because of the problems with standardization?
Neal: We thought that IEEE 802.11a would take off after 802.11b, and 802.11g would be much later. But people came out with the 802.11g chips before the standard was complete. The way that WLAN evolved was not very clean. When we went into WLAN, we put our Bluetooth on hold. But it turned out to be fine since we were able to pick up Silicon Wave and we didn't have to pay a huge amount. The economic conditions were favorable at the time. There is always a lot of opportunity in a down economy; it is a great time to invest when things are down.
MRF: The company is working in GaAs and CMOS. Has it also gotten involved in gallium nitride (GaN)?
Neal: Almost all of our business is handset related. I guess you could argue that our GPS business is not, but our main focus is on handsets. Our belief is that there is an equivalent market in dollar value in the base stations as there is in the handsets. These large-signal chips are huge for high-power amplifiers. Some of the PA die are 1 cm on a side. We believed that when we looked at what GaN had the promise to do, that the power density could be so much greater than what is out there now and that you could operate these devices at 60 to 70 V. So the breakdown voltage is very high. So we started our development in GaN, and we have made good progress. And we actually got some help from the government, a few million dollars per year, to help defray some of the development expense. The big problem with GaN has been the reliability and the surface effects and how you passivate the chip to maintain the reliability. There have been some recent breakthroughs and we feel really good where we are going with these parts for high-power amplifiers. It is always a risk when you are in new technology. We now have an infrastructure product group which is taking the components that we have made in the past and we are designing some new components, such as quadrature modulators and gain stages, and gearing them towards base-station applications. That is a diversification play for us.
MRF: We are in NYC at RFMD's first Analysts' Day, and one of the technical sessions is on wireless connectivity. As part of that topic, is the company looking at other technologies, such as ultrawideband (UWB), as a way of achieving wireless connectivity in the home for video or multimedia?
Neal: Our belief is that one of the killer aps for the handset will be voice over IP (VoIP). People pulling out their wired phones and they are going to broadband. There is no reason why we have to still have the old conventional copper twisted-pair phones. It would make sense that the cell phone that you carry with you, once it enters your home, would then work on the VoIP system, so that the one phone becomes your main phone for all purposes. Most of the young people I know only have the one cellular phone. The other killer ap is multimedia for the cell phone, and you need some way to get the data into and out of the phone quickly and we think that some of these connections will go through the network but some of them will be short-range wireless technologies. I think that UWB has a lot of promise because it offers the greatest amount of data throughput per amount of transmitted power. In a portable device, the battery is real being burdened with multimedia applications. So I think that UWB is one way to get data into and out of the phone quickly. As a company, we are looking at UWB technology, but we don't think the timing is right to develop components yet.
MRF: Can you sustain the impressive growth over the next several years?
Neal: Right now, it is a great time for us because the handset market is growing. Because we are coming from a dominant position in the handset PA marketwe are almost twice as big as our nearest competitorgoing into these transceivers, such as the Polaris chip sets, opens up huge amounts of content for us. Adding Bluetooth and GPS opens up even more content for us. We've calculated that even if the handset market remains flat for the next five years, we've got huge amounts of growth potential in the content. I like to look no further than five-year increments. The company is positioned for huge growth in the next five years. The goal is for the company to ship $1 billion per year, and we will be shipping at that rate for the next couple of quarters, so we think that the yearly goal is achievable. Once we hit that goal, we would like to take the company to $2 to $3 billion in sales annually, without making any big changes.
MRF: Is China part of that growth?
Neal: Yes it is. We have 750 people in China now that are full-time RFMD employees. Manufacturing is taking place because China is our biggest market, and they want the high-end handsets. We are also investigating design centers in Chinathere are many smart people there. We may not be sure of the timing, but the way that China is modernizing and growing, it is going to be a huge market for cellular telephones. It is going to play a strong role in our manufacturing as well. And Jazz has investments in several fabrication facilities in China.
I would like to add this. A major concern with the investment community is profitability. Our core business has been very profitable and the R & D that we are doing is considerable to grow to the next level. The overall message that we see a clear path to very profitable growth, not just growth.