Seeking Riches In 3G Chips -- Japanese IC makers aim to capitalize on the country's early adoption of this next-generation wireless technology.
Mark LaPedus
May 15, 2000, Issue: 1211
Section: Marketplace Asia
--------------------------------------------------------------------------------
Japan's cellular-phone industry is plotting a new course that could give that country's chip makers a major boost, if not a leg up, on its foreign IC competitors.
Already experiencing brisk demand, and even spot shortages, for its existing cell-phone chips, Japanese IC makers are hoping to cash in on their early adoption of third-generation (3G) wireless technology.
In fact, 3G-a digital cellular system designed to replace current wireless networks, sometimes called second-generation (2G or 2.5G) technology-seems to be a priority for many Japanese IC vendors looking to reduce their reliance on DRAMs and other commodity ICs.
"Over the last two years, we've seen a major shift from the computer to the communications age," said Gene Brannock, vice president of marketing and engineering at San Jose-based Fujitsu Compound Semiconductor Inc., the U.S. wireless- and fiber-optic-component subsidiary of Fujitsu Ltd., Tokyo. "Right now, we believe that there are some new and large opportunities in wireless, particularly in 3G."
Third-generation wireless technology is promising, but the market is fragmented and chaotic.
In theory, 3G is expected to unify the current digital cellular standards-Code Division Multiple Access (CDMA), GSM, Personal Digital Communications (PDC), and Time Division Multiple Access (TDMA). 3G will also enable handsets to obtain wireless data at speeds up to 64 Kbits/s in the initial phases, and in the future, 2 Mbits/s. Current 2G-enabled handsets can obtain data at speeds up to 14.4 Kbits/s.
To date, however, 3G is split into three competing and incompatible camps: Wideband-CDMA (W-CDMA), cdma2000, and UWC-136. Qualcomm is leading the charge for cdma2000, while Ericsson, Motorola, Nokia, NTT, and several other companies are pushing W-CDMA. The UWC-136 standard is an upgraded version of TDMA.
Despite the market fragmentation, global sales of 3G-enabled handsets are expected to grow from $1.5 billion in 2001 to $9.2 billion in 2005, according to Forward Concepts Co., Tempe, Ariz. The total investment in infrastructure equipment to support 3G cellular services is expected to grow from $1.3 billion in 2001 to $5.3 billion by 2003, Forward Concepts said.
With such numbers, Japan's chip makers aim to take the lead role in 3G. They will have an advantage, as the nation is expected to deploy this technology several years ahead of Europe and the United States, according to analysts.
By 2001, Tokyo-based NTT Mobile Communications Network Inc. (NTT DoCoMo), Japan's largest mobile operator, is expected to scrap its proprietary PDC digital cellular system and deploy the world's first 3G network based on W-CDMA.
Competitive Japanese carriers such as DDI Corp. of Tokyo are looking to upgrade their existing CDMA-based digital cellular networks to cdma2000.
But despite this fragmentation in Japan's cell-phone industry, the shift toward 3G is expected to benefit domestic consumers. 3G promises to unclog the nation's existing voice- and data-constrained airwaves, while enabling handsets to obtain voice, video, and wireless-based Internet services at speeds up to 2 Mbits/s.
Japanese OEMs are developing 3G-enabled handsets based on both the cdma2000 and W-CDMA standards, which will be sold in the domestic and international markets, said Masashi Kubota, an analyst at ING Barings, Tokyo.
"Initially, Japanese OEMs will launch their 3G-based handsets in Japan at the beginning of 2001," Kubota said. "But before that, Japan's chip makers will ship their products [to be used in 3G-based handsets] in the second half of this year."
In fact, looking to ride the coattails of Japan's early 3G deployment, the nation's chip makers hope to beat their foreign rivals to market and ship the world's first 3G-enabled components, such as multimedia chips, LCDs, LCD ICs, RF devices, and other components, to both Japanese and foreign OEMs.
Japan's IC makers could have a distinct advantage in 3G-at least on the surface. Because of Japan's thirst for small, light, and gadget-oriented handsets, the nation's cell-phone chip makers are well versed in systems-level IC design, packaging, LCD technology, and other areas.
And for years, Japanese OEMs have worked in tandem with their internal chip operations to develop the key ICs for their handsets-sometimes locking out their overseas IC rivals in the process.
Needless to say, Japan's vertically integrated companies will follow a similar pattern in 3G. "The deployment of 3G technology in Japan will give us some early design opportunities, but I don't think 3G is limited to Japan Inc.," said Andrew Burt, senior market development manager for the networking/communication business unit at Toshiba America Electronic Components Inc., Irvine, Calif.
Indeed, while Fujitsu, Hitachi, Matsushita, Mitsubishi, NEC, Toshiba, and other Japa-nese chip makers are developing various 3G-enabled components, IC suppliers from Europe, Korea, Taiwan, and the United States are also racing to ship these products.
Although it remains unclear which nation-or company-is leading the 3G-based component race, Japan's chip makers face some major hurdles in the arena.
Unlike in the past, they may not be able to develop all the necessary 3G-enabled chips for their sister handset operations or for competitive OEMs, ING Barings' Kubota said.
Battered by a long recession and unable to secure capital, the cash-strapped Japanese IC makers have been unable to develop many new, cutting-edge wireless-chip technologies, he said.
"Even though Japanese OEMs like Toshiba will try to develop everything themselves, they'll have to go outside and buy components like DSPs," Kubota said. "In terms of DSPs and [cell-phone] chipsets, Texas Instruments and Qualcomm have the lead over Japan. But in terms of LCDs and passive components, Japan will remain the global leader."
One emerging Japanese OEM has already gone outside to source some key cell-phone components. Recently, Sony Corp. jump-started its product efforts by teaming up with Texas Instruments Inc., Dallas, and Symbian, London.
Sony's next-generation cell phones will be built around Symbian's advanced software interface and TI's 2G/3G-enabled chip architecture, dubbed the Open Multimedia Application Platform (OMAP). Sony will use the chip- and software-based products to develop a line of "omnifunctional 2.5G and 3G mobile devices," said Katsumi Ihara, president of Sony Personal IT Network Co., Tokyo.
Symbian's software platform consists of an advanced operating system and an interface that supports Java, Bluetooth, and the emerging Wireless Applications Protocol (WAP). Meanwhile, TI's DSP-based OMAP platform performs the baseband functions in a handset and supports all cell-phone standards, such as 2G, 2.5G, and 3G.
Not to be outdone, Qualcomm Inc. is also pushing its 3G-enabled chipsets in Japan and elsewhere. The San Diego-based company is sampling its initial 3G-based chipsets, which support the competing cdma2000 standard.
In Japan, Toshiba is jumping on the 3G bandwagon. "On the baseband side, Toshiba has some activities in Japan," Burt said. "Toshiba is also strong in commodity products like flash memories. We also have power amps that reside in cell phones."
Additionally, Toshiba is making a strong push in multimedia chips for 3G-enabled cell phones. The company has announced what it calls the world's first single-chip MPEG-4 device for cell phones, PDAs, and related portable products.
Code named P2, the chip enables full-motion video in a portable device at speeds up to 15 frames per second. Supporting wireless data at speeds of up to 64 Kbits/s, the device integrates an MPEG-4 video encoder/decoder, a speech codec, an audio/video multiplexer, and a 16-Mbit embedded DRAM. The device will begin sampling in the third quarter of 2000.
"I don't believe that the current generation of phones will adopt MPEG-4," Burt said. "The technology is geared for next-generation phones."
The chip is drawing interest from OEMs in Japan and Korea. U.S. OEMs are taking a wait-and-see approach, however.
"Japan and Korea are ahead in terms of developing multimedia phones with MPEG-4 capabilities," Burt said. "Japan is especially aggressive, but it could be more of a cultural thing than anything else. Japanese consumers like small, feature-packed gadgets, but that doesn't seem to be a real concern for consumers in the United States."
In addition to multi- media chips, Japan's IC makers are making a bid to compete in other 3G-enabled component markets such as RF devices.
Earlier this year, Mitsubishi Electric Corp. and Analog Devices Inc., Norwood, Mass., announced plans to jointly develop direct-conversion RF chips for W-CDMA-enabled cell phones.
Based on ADI's technology, the device includes variable-gain amplifiers, baseband channel filters, and a wide-dynamic-range logarithmic amplifier for RSSI (received signal-strength indication) detection.
The device streamlines handset designs by enabling incoming high-frequency signals to be sent directly to baseband, without the need for additional conversion circuitry such as oscillators and filters.
The chip, to be shipped later this year, also reduces the cost and bill of materials for next-generation handsets, according to Kenji Itoh, section manager of Mitsubishi's next-generation radio-terminal development department.
Although Mitsubishi appears to have a jump on 3G-enabled RF technology, Japan's chip makers generally are lagging in this segment.
"I would not characterize [Japanese chip makers] as being on the cutting edge in RF," said Mark Burkett, senior vice president of Santa Clara, Calif.-based California Eastern Laboratories Inc. (CEL), which designs and distributes NEC Corp.'s RF chip products in the United States. "U.S. companies like Conexant, Anadigics, and RF Micro have been more aggressive in developing new technologies."
For example, Conexant Systems Inc., Newport Beach, Calif., has announced an RF-based chipset for the cdma2000 standard, while RF Micro Devices Inc., Greensboro, N.C., has rolled out a similar device for W-CDMA applications.
Generally, Japanese cell-phone chip makers are not known for highly integrated, RF chipset products. Rather, they are powerhouses in discrete components, particularly those based on bipolar and gallium arsenide (GaAs) process technologies.
Discrete devices are still critical components for handsets. Pressured to reduce their costs and design cycles, OEMs continue to rely on these flexible discretes to get their products out more rapidly.
In fact, some discrete products are hard to come by. "When the cell-phone market took off last summer, it took us a little by surprise," CEL's Burkett said. "As a result, our products are on allocation right now. Our lead times are being pushed for our LNAs [low-noise amplifiers] and VCOs [voltage-controlled oscillators]."
In response, NEC is expanding its gallium arsenide and bipolar wafer-fabrication facilities in Japan. Currently, the company has a 5-in., GaAs-based fab near Kyoto that makes devices based on an advanced GaAs process technology called heterojunction bipolar transistor (HBT).
NEC also prides itself in developing ICs and packages for small-form-factor products. "We're strong in terms of developing small-size discrete products such as LNAs, VCOs, and power amps,'' Burkett said. "NEC's also strong in manufacturing and packaging."
Recently, NEC developed what it claims is the world's smallest package for transistor-level products. With an overall footprint of 1 x 0.6 x 0.55 mm, the new M23 package is ideal for use in cell phones and other portable packages.
Like NEC, Fujitsu is preparing for the 3G wireless explosion by expanding its manufacturing capacity, developing new process technologies, and beefing up its product portfolio.
Fujitsu Compound is developing discrete-oriented ICs for 3G wireless applications, particularly power amps for handsets and base stations, Brannock said.
But the company is not experiencing availability problems with its cell-phone chips-at least for now. "We're not seeing any shortages," Brannock said. "There are other components that are a limiting factor."
The Fujitsu executive was referring to the current worldwide shortage of select cell-phone components-SAW filters, LCD driver ICs, and some RF devices-which has hampered product shipments for many OEMs this year.
Still, Fujitsu Compound is expanding its 4-in., GaAs fab in Japan, which makes devices based on a pair of technologies-metal semiconductor field-effect transistor (MESFET) and pseudomorphic high-electron mobility transistor (PHEMT). The company is also developing an HBT process.
Although Hitachi Ltd. has not disclosed its 3G chip plans, the company is making a major push in several new wireless-IC markets such as Bluetooth and CDMA-based RF devices. The Tokyo-based company is already a powerhouse in GSM-enabled baseband chips and power amplifiers.
"We're not just a DRAM or SH [microprocessor] house," said Peter Clark, president and chief executive of Hitachi Semiconductor (America) Inc., San Jose.
Hitachi is sampling an RF device that supports dual-band GSM handsets and the General Packet Radio System (GPRS) standard. The HD155131TF integrated transmit/receive chip supports the 900-MHz and 1.9-GHz frequency bands.
The forthcoming device incorporates dual-phase-locked-loop, LNA, and other functions. In addition, an offset phase-locked-loop system is used for frequency conversion on the transmitting side, thereby eliminating the need for a discrete antenna block.
The HD155131TF uses a 0.35-micron BiCMOS process and integrates most of the RF block functions in a single chip using silicon-on-insulator technology and a deep-groove separation structure.
---
Eyeing 3G market's potential
Analysts expect Japan to deploy 3G technology ahead of other regions, and Japanese chip makers hope to cash in on the country's early adoption of 3G technology.
3G is expected to unify current wireless standards-CDMA, GSM, PDC, and TDMA.
The technology is expected to enable handsets to obtain wireless data at speeds up to 64 Kbits/s in the initial phases and, in the future, 2 Mbits/s.
Global sales of 3G-enabled handsets are expected to grow from $1.5 billion in 2001 to $9.2 billion in 2005, according to Forward Concepts Co.
It's uncertain who will lead the 3G-component race.
Cash-strapped Japanese IC makers have been unable to develop some new wireless-chip technologies.
Japanese OEMs like Toshiba will try to develop the chips they need themselves, but analysts say they will have to go outside and buy components like DSPs.
ebnonline.com
Copyright © 2000 CMP Media Inc |
