MRF: TriQuint has been such a key part of the development of gallium arsenide (GaAs) in this industry. It must be fascinating for you to look back at the company's role during your time there.
Peters: It is truly amazing to contemplate the growth of both our company and the industry overall. In 1985, GaAs was "the" big news in RF and microwave semiconductor technology. The mood and some of the projections for extravagant growth were thrilling. GaAs is one of those rare technologies that actually lived up to the initial hype. It has obviously enabled the capabilities of cellular phones, a market that has been critical to TriQuint's success, but it's interesting to see GaAs growth opportunities in a multitude of new applications, many of which simply didn't exist until recently.
MRF: When TriQuint was first founded, transmit/receive modules for phasedarray radars were being touted as the "killer application" that would drive the volume of GaAs devices.
Peters: I remember when phased-array radar was seen as the case study for potential application for GaAs FETs and MMICs because the largest arrays require many devicesit became a great example of potential realized. Today, GaAs electronically steered radar systems have come into their own and are critical elements of any integrated defensive or targeting platform. TriQuint devices are aboard F-16, F-18, F-22, and B2 aircraft and retrofits, as well as the newest F-35 Lightning, the Joint Strike Fighter (JSF). Phased-arrays play a critical role in the EQ-36 battlefield radar systems that are helping to defend allied forces in Iraq and Afghanistan. But in terms of sheer volume, the number of GaAs devices produced for mobile phones in a single year probably eclipses the total number used in phased-array radar systems since day one!
MRF: How did TriQuint actually take shape?
Peters: Tektronix saw a requirement for technologies that could drive the performance of its high-performance test equipment and invested heavily in GaAs. It established a research group within the company that was the foundation of TriQuint. Tektronix spun TriQuint off as a research organization with the hope that it could support its internal needs and recoup some of its costs.
MRF: Obviously, TriQuint's prospects grew beyond those of what Tektronix might need.
Peters: They did indeed. We immediately searched for foundry customers and markets beyond test equipment where GaAs technology could provide technological benefits. One of those markets was GPS and we worked with Magellan to support their growing portfolio. We also received contracts from Motorola, DARPA, and Bell Northern Research. In 1989, we were the first organization to employ plastic packaged GaAs devices, which produced significant cost benefits. Also, in cooperation with Nortel, TriQuint was launched into the long-haul telecommunication sector within Nortel's multiplexers and demultiplexers. In fact, the capabilities of our products actually had a great deal to do with Sprint's ability to claim that its quality of voice service was so good you could hear a pin drop.
MRF: What else propelled the company forward?
Peters: In the early 1990s, Tektronix wanted to recoup its investment in TriQuint. We ultimately merged with Gazelle Microcircuits, a foundry client, and Gigabit Logic, which competed with us in digital products, to provide greater technological breadth, which effectively ended our history as a pure R&D organization. The new opportunities provided by the merger allowed us to pursue ASICs to complement the foundry part of our business. It's important to note that in this period, the cellular industry was just beginning to take shape.
The year 1994 proved to be both a demanding and exciting one for us, as we acquired AT&T's HEMT process, created our first enterprise-wide strategic plan, and received ISO 9001 certification. At this time, we were working closely with Nortel, our largest customer, and one of our products experienced a failure. We quickly identified and corrected the problem. Our approach to problem solving impressed Nortel, and cemented a strong relationship. The knowledge we gained from the process also turned into a significant competitive advantage, which we closely held and applied to future process technologies. We made lemonade with the lemons.
MRF: By that time, the cellular industry had taken off, which must have been a watershed event in the company's history.
Peters: That's definitely an understatement. Mobile phones were another application in which GaAs had a clear advantage over its silicon counterparts, and we rapidly increased our production of devices for these products to the point where we were severely straining our production capabilities. Our plan was to move to a new site and establish 6-in. wafer capability. In 1997 we achieved both goals while also intensifying our MMIC development efforts. Texas Instruments had decided to get out of the GaAs business, and we merged their capabilities with our own, giving us more customers, leading-edge technologies, and some terrific design engineers. Interestingly enough, TI had sold the GaAs MMIC business unit to Raytheon, but the government stopped the sale because it felt Raytheon would then have a monopoly in the GaAs MMIC defense and aerospace business segment. We were able to buy this GaAs business at a great price.
MRF: At the tail end of the 1990s, what did the market potential look like for TriQuint?
Peters: At the end of the decade we were experiencing greater than 25-percent growth across all business divisions but the landscape was rapidly changing. The wireless market continued exponential growth, while the telecom sector that was our initial and continuing core business, was declining. Our yields were the best in the industry and we introduced 50 new products in a single year, and receivers for Nokia became a very important product for us, and more than 200 million were produced. In addition, the HBT process development initiated in Texas resulted in technology transfer to Hillsboro, OR and boosted our production volumes.
MRF: How did TriQuint fare during the "dark days" of the early 2000s? Peters: We felt the downturn like everyone else and it was a trying time. But 2001 was still a watershed year as we merged with Sawtek, a major manufacturer of surface acoustic wave (SAW) devices. We also established a sales presence in China. We followed this the next year with the acquisition of Infineon's GaAs unit and Agere's optoelectronic operations. Ralph Quinsey was named President and Chief Executive Officer, a position he holds today, and his vision and dedication have proven to be invaluable to the company's growth, commitment to quality, and corporate culture. In 2003 we began pHEMT development in Hillsboro and created our first modules and power amplifiers for mobile phones. At this point, we had a very formidable array of products to offer our customers, from multiple GaAs foundry processes through a wide range of MMIC devices.
MRF: Since those watershed years, the company has grown to serve many more markets besides mobile phones and has continued on the acquisition trail.
Peters: Both true. We acquired TFR Technologies, which added bulk acoustic wave (BAW) expertise to our SAW capabilities, and then Peak Devices, which expanded our networks product line with new broadband, high efficiency RF power technology that we call Power- Band. We also acquired WJ Communications, formerly Watkins Johnson Company, adding further device technology and customers to our arsenal. Our latest acquisition, TriAccess Technologies, gave us much greater capabilities in the massive cable market, with proprietary on-chip linearization techniques that they apply for traditional hybrid fiber coax cable systems as well as fiber to the home (FTTH). Our unified product line serves the full RF needs of the CATV industry, which is growing globally at a very fast pace.
Continue on page 2
MRF: Now that you have so broadly expanded your capabilities and markets, what do you see for the future of both the company and the marketplace? Peters: All signs point toward continued growth of the cellular marketplace, especially as carriers roll out LTE in 700 MHz and both LTE and WiMAX in the 2.3 to 2.4 GHz and 2.5 to 2.7 GHz bands. Networks for 700 MHz require user equipment to cover this band that simply does not exist today and the nascent WiMAX and LTE bands above 2 GHz represent additional build out opportunity. This represents terrific upside for TriQuint and many other companies as well. Regardless of how the public safety wireless initiative plays out, it will also be another major market for our products.
MRF: The issue of backhaul has taken a back seat to other network issues until recently. Do you see this situation taking on greater importance?
Peters: Backhaul is an enormous challenge for carriers, equal to the issue of how to handle over-the-air network traffic, which has been getting most of the attention. Backhaul is already a bottleneck and will be even more onerous when LTE and its next generation, LTE-Advanced, are deployed. Backhaul, which is the transmission path from cell sites to network hubs, is served almost exclusively in the US today by leased copper T1 lines; microwave links and fiber have much greater penetration in other parts of the world. Carriers are feverishly exploring alternatives, which essentially means newer and higher speed fiber and microwave links, since no other technologies can provide the required throughput.
MRF: What is the solution in the United States?
Peters: The massive bandwidth required by fourth-generation networks and the transition from Time Division Multiplex (TDM) to Internet Protocol (IP) transmission is likely to relegate copperbased backhaul to the history books. Even today, levels of traffic are severely stressing networks and their current backhaul capabilities. In relatively short order, leased lines will be replaced by microwave and millimeter-wave links and fiber-optic networks. TriQuint has a rapidly growing line of products for both of these backhaul solutions, whether the technology is a microwave or millimeter-wave point-to-point radio or a lightwave system.
MRF: How about the future of cable systems?
Peters: The cable industry is undergoing a massive upgrade to the DOCSIS 3.0 standard that will enable higher downstream and upstream speeds. With our acquisition of TriAccess Technologies, a former foundry services customer, we now have a broad product portfolio to serve all portionsheadend, network infrastructure, and customer premises equipmentof this rapidly growing market.
MRF: Earlier, you mentioned seeing new opportunities for GaAs. Please elaborate.
Peters: Communications are moving from voice, to data, to video. New applications are emerging such as automotive radar, smart meters, millimeter-wave imaging and tissue ablation in medical procedures. These communications and emerging new applications will require GaAs transmit and receive devices, and the frequencies at which they operate will continue to push well into the higher reaches of the millimeter-wave region. Our GaAs technology portfolio makes it easier and more cost-effective than ever to achieve high-performance with low power consumption.
MRF: How does TriQuint address the defense market?
Peters: TriQuint is proud to be a leading supplier to the defense and aerospace industry. You will find our products in virtually every type of defense system on terrestrial, airborne, and sea-based platforms. We offer modules, packaged products, as well as foundry services to defense suppliers. We are one of only a handful of DoD-Accredited 1A Trusted Foundry' service providers. We have R&D contracts including several gallium nitride programs with various government agencies including the US Air Force Research Laboratories (AFRL) to develop new GaN modules for unmanned aerial vehicles and DARPA to create complex, high dynamic range GaN circuits for future defense and aerospace applications. We can package die-level devices and can help customers with on-chip or multichip module development all in-house, which especially appeals to securityminded defense customers.
MRF: Gallium nitride (GaN) is of great interest for defense applications. What is TriQuint's GaN roadmap?
Peters: Just a few years ago GaN was considered only developmental, and now it's being used in critical battlefield communications and jamming systems. TriQuint has been a leader in the development of GaN for many years, and has secured many contracts to design, develop, and build GaN products for use in defense and commercial applications. We offer GaN foundry services as well as standard products. We think GaAs and GaN will peacefully co-exist for the foreseeable future, each serving their unique markets.
MRF: In short, what does your crystal ball show for the next few years? Peters: There are more applications for wireless connectivity and easier, lessexpensive ways to provide it than ever before. Cellular user equipment is just the most obvious example, but far from the only one. As the world untethers itself from wired connections and countries bring voice and data connectivity to their populations, we anticipate strong demand for the products with which we have established our leading position. I feel very confident that TriQuint, as well as many other companies in our industry, will experience strong growth in the coming years. I think it's an exciting time to be part of it.