Wireless and broadband: Opportunities for chipmakers
05/01/2001
Communications applications once constituted the ugly brother of the PC world in terms of its relationship to the growth of the semiconductor market. These systems were generated in relatively low volume and represented minimal profit opportunity/device. After all, the semiconductor dollar content/PC can easily exceed $500/device, and unit shipments, which have been in the tens of millions of units/year, have now surpassed the one hundred million units/year mark.
 |
Growth in the communications market, however, has elevated the total available market (TAM) for semiconductors used in such devices to well beyond the tens of billions of dollars mark. According to the Semiconductor Industry Association, the segment is now recognized as a key driver of the overall semiconductor industry, a significant acknowledgment, since the SIA virtually ignored this sector a few years ago. Two main trends now drive the communications marketplace: mobile access à la cellular phones and the quest for broadband capabilities. Not surprisingly, wireless and broadband are converging in both fixed and mobile spaces.
Cellular phone technology is one of the areas in which this convergence is most prevalent. By most accounts, cellular shipments reached about 415 million units in 2000, and the total TAM available to producers of transceiver-related semiconductors was >$10 billion (see table).
 |
Growth in the mobile wireless market continues to be fueled by the increasing number of cellular phones in established nations; their expanding footprint in developing nations; and a very high replacement rate. The latter is particularly important to handset, and hence, semiconductor manufacturers, since high replacement rates equal higher revenues. This trend is also expected to continue with the rollout of more advanced third-generation services and even the inclusion of two-and-a-half-generation capabilities. Devices for mobile cellular technology will require a significantly richer feature set than those for narrowband communications, a factor that should propel shipments and replacement rates upwards. In addition, some of these devices will increase the semiconductor value content/phone from $25 to $35 for basic digital phones to upwards of $100 for more complicated devices in the short-term. Data-enhanced (two-and-a-half and third-generation) communicators are expected to account for ~70% of the total cellular phone market by 2005.
Meanwhile, fixed wireless broadband applications (often referred to as broadband wireless access or BWA) are showing considerable potential as well. Both incumbent and competitive local providers are interested in these systems, either to extend the reach of an existing infrastructure or to establish competitive local services. A wide range of frequencies is being considered and/or implemented for these applications, but generally, they fall into two main categories, below and above 11GHz. Volume shipments of these applications have not yet begun, but collectively, BWA platforms are expected to establish a significant footprint throughout Europe, North America, Japan, Korea, and parts of Latin America and the Middle East. Overall, BWA technologies are expected to account for 20% of all broadband wireless subscribers by 2005. For semiconductor producers, the potential TAM is significant.
 |
Semiconductor suppliers to the wireless world are numerous and often highly specialized. Production of a wireless system requires expertise in mixed signal, analog, digital, and RF or microwave chip design. As a result, suppliers have traditionally focused on specific functional areas of the transceiver. The RF realm is dominated by such small firms as RF Microdevices, Triquint Semiconductor, Anadigics, Celeritek, and Alpha Industries.
Meanwhile, larger companies tend to focus on a systems perspective that centers around IF, mixed signal, and digital aspects of the radio chain. This latter group includes vendors like Philips, Lucent, Infineon, Analog Devices, and QUALCOMM. A few manufacturers like Conexant and Motorola are producing entire wireless semiconductor systems internally. Specialization and the resulting market fragmentation is expected to continue in the wireless world, since tier one players use several semiconductor manufacturers to optimize their systems' performance. The ability to produce an entire wireless semiconductor system, however, is becoming more important as firms with limited wireless design experience compete in the rapidly expanding wireless application market.
Though BWA and mobile devices are poised to represent two of the most exciting developments in the wireless world, they are far from the only ones. Wireless local area networks (WLANs) continue to forge ahead in both consumer and commercial sectors, and global positioning system (GPS) applications are demonstrating double-digit growth as well. The WLAN semiconductor market is led by Intersil, which has been able to garner a significant lead in market share by providing complete and low-cost solutions that center around its Prism designs. Leaders in the GPS field include SiRF, Motorola, Conexant, and Trimble.
Beyond the wireless world, the communications market is also being shaped by broadband capabilities in the wired realm. Digital subscriber line (DSL) and cable modem technologies are being rolled out to address corporate and consumer demand for fast data and Internet capabilities. Total DSL modem shipments, customer and central office side, are expected to exceed 15 million by year-end 2003. Meanwhile, shipments of cable modems are expected to exceed the 10-million-unit mark in 2002. Total semiconductor opportunity for these two devices, central office and customer modem, is expected to pass the $1 billion mark by year-end 2002.
The dynamics in these markets are significantly different from those in the wireless realm, despite the fact that cable operates within RF frequencies. Integration is easier to establish because communications systems' operating frequencies are relatively low, and analog design, although still important, is not as big a challenge. In fact, many would argue that the most challenging aspect of designing DSL and cable modem chip sets is in media access controller and/or digital-signal-processing functions. Thus, relatively large system or subsystem houses such as Broadcom and Analog Devices dominate this market. In addition, there are few specialized analog chip designers that focus on these markets as is the case with wireless technology.
Penetration rates for broadband services, mobile and fixed, are minimal today. Of the roughly 320 million Internet users around the world, fewer than 10% are using broadband delivery technologies. DSL, cable, and wireless will be employed to accommodate these users, and we can expect total penetration of broadband services to exceed 80% over the next 5-10 years. Meanwhile, total Internet use will have grown to 796 million by year-end 2005. But mobile telephony, perhaps, is likely to remain the most appealing market for semiconductor suppliers. Due to a high growth rate and a significant replacement rate, these devices are expected to provide growth to chipmakers for some time. The convergence of computing with wireless mobility will provide a significant boost to both the wireless and overall semiconductor market for years to come.
Andy Fuertes is VP of communications technologies research at Allied Business Intelligence Inc., 202 Townsend Square, Oyster Bay, NY 11771; ph 516/624-3113, fax 516/624-3115, www.alliedworld.com.