Sierra Monolithics, based in Redondo Beach, Calif., was founded in 1988 by Charles Harper, today the company's chairman; Binneg Lao, chief scientist; and David Rowe, chief technology officer. Initially, Sierra Monolithics designed hardware, based on the off-the-shelf chips, for the military and aerospace communities. The products, still in production, included radar beacon transponders for friendly military forces and the video/data transmission system for the Predator, an unmanned aircraft used by the U.S. Air Force for surveillance and combat. In the 1990s, Sierra Monolithics started designing its ICs for design services contracts for companies such as IBM, Research in Motion and American Microcircuits.

The company's core expertise is in mixed signal design. In 2001, Sierra Monolithics (SMI) made a strategic move to capitalize on this expertise by transitioning from a design company to a product company, designing its own products for sale in lieu of designing products for other companies. Then it established a plan to enter the long-haul networking market. Financed by venture capital from Storm Ventures and US Venture Partners, the company developed 40 gigabit-per-second "serializers" and "deserializers" to support data transmission on fiber optic networks.

Today SMI is the primary supplier of these products for OC-768 core network transponders. Two years ago, SMI entered another new market when it began developing an RF chipset for WiMax. The company is now shipping RFICs [Radio Frequency Integrated Chip] to the leading OEMs in the fixed WiMAX market, including its current chip, SMI7035, that supports the 2.3 to 2.7 gigahertz and the 3.3 to 3.8 gigahertz bands. The SMI7035 is compatible with Intel's Rosedale IC as well as several other manufacturers that supply the PHY/MAC for 802.16-2004 applications.

WiMax.com: Can you explain the differences in the design approach, manufacturing approach and most importantly the interface elements between an RF chipset type of silicon product and the so called "baseband" or media access controller (MAC) and physical layer (PHY) companies such as Intel and Fujitsu?

SMI: There are three basic pieces to a digital radio, whose function is to transmit bits wirelessly. The RF section picks up signals on an antenna, amplifies them and converts the frequency so that the signals can be processed by the radio's physical layer (PHY), which converts the signals into digital form. Then the radio's media access controller (MAC) controls the frequencies used by the radio and coordinates user time slots based on activity at a base station.

Sierra Monolithics makes the RF transmitter and receiver in a single transceiver chip that is called a transceiver RFIC. The design of the transceiver employs microwave, analog and digital design techniques and silicon germanium technology for fabrication. Other companies make the MAC and PHY layers, which comprise the baseband section in a single chip or dual chip implementation. The baseband is typically fabricated in CMOS [Complimentary Metal Oxide Semiconductor] technology. The signal is passed between the RFIC and the baseband and control signals are routed from the baseband processor to the RFIC.

WiMax.com: Who uses your WiMAX RF IC products? What types of manufacturers are they?

SMI: The products are bought by OEMs of WiMax equipment, including Siemens, Alvarion and HuaWei, ZTE, Aperto and GemTek. Other companies also make assemblies for them or customer premise equipment systems sold through electronic retailers.

WiMax.com: Who are some other RF chipset companies currently building WiMAX products and what is the biggest difference between Sierra and those firms?

SMI: Other makers of RF chipsets are startups RF Magic and SiGe Semiconductor and established chipmaker Texas Instruments The major difference between Sierra Monolithics and these companies is that it uses one integrated WiMax RF chip, as opposed to two or three, making it more economical. A good product at the lowest possible cost is critical to sparking and expanding the WiMax market.

WiMax.com: Do you provide reference designs for your customers and can you describe what is included in a reference design at a high level?

SMI: Yes. Sierra Monolithics provides prospective customers with evaluation boards encompassing a complete RF transceiver, interfacing on one end to an antenna and on the other to a baseband chip. The board includes the SMI 7035 chip, a power amplifier, a low noise amplifier, a transmit/receive switch, voltage regulators and few other integrated circuits for support. Several baseband companies ship our evaluation boards along with their baseband boards to demonstrate total system capability.

WiMax.com: What initial spectrum ranges will your RF chipsets support and will they be only for 802.16-2004 Fixed WiMAX products or can you readily shift your designs for 802.16e boards?

SMI: The SMI7035 currently supports the 2.3 to 2.7 gigahertz and 3.3 to 3.8 gigahertz spectrums, representing most of the frequencies now in use globally. In design is another RF chip for the 5.1 to 5.9 gigahertz band, which will serve an emerging WiMax frequency allocation that will eventually be used in the United States and other countries. The chip will go into beta test during the first quarter of 2006. Some of Sierra Monolithic's designs for the fixed WiMax market will carry over to the 802.16e mobile WiMax market, but the company has to develop a new chip for the latter market. Sierra Monolithics is currently talking to several handset equipment makers in Japan, China and Korea that are considering using the wireless RF chip in tandem with a baseband chip supplier

5.8 GHz WiMAX Transceiver Evaluation Board

WiMax.com: Will you need to deliver separate chips for Fixed WiMAX and Mobile WiMAX? If so, how much difference is there in the two designs and in the actual chips?

SMI: Yes. There are significant similarities and differences in the two chips. Several of the major building blocks of intellectual property developed by SMI for its fixed WiMAX RFIC, as well as other projects, form the basis for the mobile IC. But a mobile chip must be very low power. And the interface to the baseband chip must be digital, as opposed to IF, the interface used in our current chip. SMI will develop a separate mobile WiMax RFIC, based on the foundation of our existing RFIC.

WiMax.com: Can your chips be easily tuned for other spectrum ranges such as 2.4 GHz and the US 3.65 GHz unlicensed bands if the Forum chose to develop product profiles for those spectrum ranges? Also, would the economies of scale for licensed band chipsets translate well to chips tuned in these ranges?

SMI: Yes. SMI's current product supports these bands. The SMI7035 already covers about 75% of spectrum ranges globally. With the addition of our 5 gigahertz band chip, Sierra Monolithics will cover every licensed and unlicensed spectrum range globally for the foreseeable future. A single chip solution that can be tuned to any frequency, licensed or unlicensed, creates significant economies of scale.

WiMax.com: Do you see WiMAX being disruptive to 3G technologies or supportive---and why?

SMI: WiMax is complementary to 3G. 3G supports reasonably high quality voice and digital communications. But for high quality real-time video, users need the much faster data stream offered by WiMax. WiMax, unlike 3G, can also support voice over IP. Ultimately, there is a good chance that one mobile phone will encompass both technologies and use the one that makes most sense, depending on location, class of service and other variables. SMI's view is that WiMAX will not supplant 3G anytime soon because 3G networks have already been built globally and have proven their reliability. So, initially, WiMax will support 3G for users who want extra features available in certain locations.

WiMax.com: In practical terms how does your technology reinforce the WiMAX business case of your customers and ultimately wireless providers, both fixed and mobile? Is it just cost savings or are there technical details that really matter?

SMI: Sierra Monolithics reinforces the business model for WiMax in two ways. A single-chip transceiver helps push down costs, enlarging the market. Secondly, the high performance of the radio enables more bits to be transmitted through the same amount of spectrum. New technologies like WiMax invariably succeed on the strength of their ability to offer popular capabilities at low prices.

WiMax.com: Does your firm build RF chipsets for other purposes? Is Sierra a purely WiMAX oriented chip vendor? If so or largely so, why are you betting so much of the company on WiMAX---after all it has very credible competition?

SMI: Sierra Monolithics isn't betting the company on WiMax, although it is an important part of our future. SMI also has revenue streams from several other markets unrelated to broadband wireless access. As previously noted, for example, SMI is the leading supplier of SERDES products for the OC-768 optical telecom market. These are critical, high value mixed signal products used in routers and transponders for core network applications. SMI also has a significant U.S. military business. In addition, R&D is underway for next-generation automobile radar technology and jet fighter radar systems. The company is pursing WiMax because it is a good match for our core RF and mixed signal capability, one with significant growth potential.

WiMax.com: Where do you see the future of WiMAX and Sierra Monolithics leading? Do you believe WiMAX will be a huge driver for your business?

SMI: Sierra Monolithics has invested nearly $6 million in R&D and manufacturing on WiMax and is very bullish about its prospects. It's an excellent standard, developed by a vast body of engineers, and it is very flexible in terms of data rates and network optimization. It maximizes the strengths of current technology by packing more bits per second into available bandwidth than any alternative and accommodates many functions on the same network. Fixed WiMax interoperability trials are proceeding well globally. We are bullish about WiMax prospects and our role as a major WiMax player.

Vivek S. Pendharkar
Interim-CEO

Vivek S. Pendharkar brings over 20 years of experience in the semiconductor and computer systems industry. Most recently, Vivek was Vice-President at Cypress Semiconductor Corporation. Vivek joined Cypress in July 2001, as part of the acquisition of Lara Networks, where he was Vice-President of Engineering. Prior to that Vivek worked at Intel as Director of Engineering on the IA-64 program that developed the first 64-bit Itanium Processor. Vivek joined Intel from Hewlett-Packard, where he worked for 12 years in various engineering & management positions in the Computer Systems Group that developed PA-RISC processors and systems.

Vivek holds BSEE from BITS Pilani, India; MSEE from Virginia Tech; and has also attended Graduate School of Business at Stanford University (SEP 95).