More
    HomeMobile EuropeA process of power

    A process of power

    -

    Tony Dennis looks at the impact of chipsets on handset supplies in Europe, and examines what kind of influence they might have on future handset development

    With the increasingly more complex nature of today’s mobile handsets — especially those being sold in Europe — the supply of chips and chipsets is playing an ever-more important role for handset vendors and their network operator customers. In many cases, poor consumer acceptance of specific handsets — due to low battery life and heavy, bulky casings — can frequently be blamed on the chipsets they are utilising. Furthermore, delays in shipping certain smartphone and high-end feature phones have been blamed directly on component chip shortages.

    A classic instance of chip shortages occurred back in December 2003 when Tom Lynch, president for  Motorola’s PCS division was forced to admit that:  “[Shipping] unit volumes are lower than our customers are asking for due to supply constraints for the integrated-camera components.” His colleague, Bob Perez, general manager for supply chain operations with Motorola PCS, claimed his company was not alone in its suffering.  “The supply constraint for integrated-camera components is an industry-wide problem,” Perez maintained. “Motorola is feeling a larger impact because of the extremely limited supply of the smaller camera technology that we use for our handsets, which enables a compelling clamshell form factor and stylish design.” 

    Some observers casts doubt on whether the problem was genuinely industry-wide, though. Paolo Pescatore, a senior research analyst for EMEA wireless with IDC, claimed that it was, in fact, quite rare for handset manufacturers to be hit by chip shortages.  “It’s more a question of [the handset vendor] underestimating its own resources and capabilities plus poor management,” Pescatore told Mobile Europe. “Vendors know when there is going to be a blip in user demand — around Christmas — and they can easily plan ahead for it.”

    The general consensus, however, is that a shortage of camera lenses and associated digital optical components is an unusual experience for the handset industry in general. By contrast, low battery life and larger physical dimensions for handsets are blamed directly on component choice and this applies in particular to 3G handsets.

    The solution, according to Michael Civiello, vp for marketing and business development with specialist chip supplier, Zyray Wireless, is to employ a separate, dedicated 3G chip. Civiello estimates that in 2004 there will be a market for around 10-13 million W-CDMA handsets in Europe and that 95 per cent of the problems with 3G will have been ironed out. “Motorola and NEC’s 3G software and hardware was set in stone something like two and a half years ago,” Civiello claimed. “We’ve learnt from their pains and experiences. Our 3G chips are up-to-the minute and include all the latest advances. This enables handset vendors to snap on a reliable, low cost 3G solution to an existing design.” Describing Zyray’s chips as “Velcro-like”, Civiello was forced to admit that presently its products only bolt onto Infineon based handsets but claimed that still meant he was targeting 15 per cent of the total handset market.

    Overwhelmed with demand

    An area where Zyray may well enjoy an advantage concerns EDGE [Enhanced Data rates for Global Evolution]. He claims that a quick straw poll of Zyray’s potential clients showed that they are currently overwhelmed with demand for a handset which supports both W-CDMA and EDGE. “I’ve yet to see a handset shipping which offers both UMTS and EDGE. However, I fully expect to show a reference design for such a handset at 3GSM 2004 which is aimed at potential customers in both Korea and Europe,” Civiello commented.

    Despite EDGE’s original public profile as purely an American technology, to date eight European network operators have announced their intention to offer EDGE over their networks including Bouygues Telecom in France. A strong appeal for operators is the possibility of a tri-mode handset which can drop down from 3G to EDGE’s enhanced data speeds rather than falling immediately back to GPRS.

    Another very vocal advocate of EDGE is handset reference supplier, TTPCom. The company has established strong relationships over EDGE technology with Intel and ADI for baseband chips and with Renesas for the necessary radio chips. The company argues that the real costs for operators lie not with infrastructure roll out but with proving that such handsets work over their own networks. For that reason TTPCom has been co-operating closely with Rohde & Schwarz and has successfully completed 70 test cases already. Consequently TTPCom confidently predicts that by mid 2004, handsets utilising its EDGE designs and technology will be shipping.

    Many of TTPCom’s customers are Asian based manufacturers and TTPCom argues that this is one clear case where technology provides a market advantage. Presently there is only one EDGE handset shipping — Nokia’s 6200 — and no sign that the other Tier One handset vendors are ready to compete. So, just as NEC has exploited the fact that out of the top vendors only Motorola currently offers a competitive 3G handset, so other Asian manufacturers will seize the opportunity offered by EDGE, TTPCom believes.

    Mainly thanks to its long established relationship with Nokia, Texas Instruments and its OMAP chips in particular have long dominated the handset market. Sales figures for 2003 have been very encouraging. According to the latest report from Forward Concepts — which monitors sales of DSP chips in particular: “After reviewing the Q4 market guidance of market leader, Texas Instruments, and increased projections of cellular phone shipments by the financial community, we have increased our earlier 15 per cent (DSP chip) shipment growth to 20 per cent for 2003. Also, after earlier cutting back our 2004 forecast, we are also raising it from 20 per cent to 25 per cent. [That’s impressive] when you compare this with the 2.3 per cent increase of the [total chip] market over the same period – DSP is a clear winner.”

    However, efforts by the world’s leading chipset vendor, Intel, to crack this market are still failing. “While many of the leading handheld vendors (H-P, Dell, Toshiba, etc) use Intel processors, the company has not yet made any impact on the smartphone segment, ” said Mike Welch, an analyst with  market watchers, Canalys. “The majority of even the small percentage of smartphones that are running a Microsoft OS [Operating System] are currently using TI chips.”

    Nevertheless, as the importance of high-end feature phones and smartphones grow, so does the importance of processing power. “We’ve already seen the effects that the demand for more processing power can have in handsets. We’ve seen handsets that crash as a result. The demand for greater processing power is only going to get stronger,” argued Paolo Pescatore. “Handsets now need MMS capabilities; they need built-in cameras; they need to support Java and to run games. That’s obviously going to create a demand for faster processors and handsets that can truly multi-task  — like PCs can. Users are impatient and they aren’t going to want a handset which forces you to wait before you can make a call.”

    Which is perhaps why Intel thinks it might have an edge with its forthcoming chip, code-named Bulverde. It’s capable of running a game originally written for the Microsoft X-box; plus it features ‘QuickCapture’ technology enables the user to take 4 Megapixel photos and capture video at 40 frames per second. Additionally the Bulverde will offer ‘Wireless Speedstep’ technology which dynamically adjusts the power and performance of the processor depending on the applications running. Intel expects to deliver the Bulverde chip to customers this year (2004).

    Another major challenge to Texas Instruments is expected to come from StarCore, LLC, a new company focused on DSP technologies formed out of the StarCore joint venture between Agere Systems and Motorola and joined by Infineon Technologies. According to Thomas Lantzsch, CEO with StarCore: “We aim to fundamentally change the competitive playing field through licensing and wide availability of DSP cores. That will help manufacturers…deliver higher levels of performance and miniaturization while accelerating time to market and lowering overall production costs.” According to Will Strauss, Forward Concepts’ president: “Backed by three leaders in DSP and communications, the new company has a great opportunity to proliferate the StarCore architectures to a broad cross section of semiconductor manufacturers, OEMs [Original Equipment Manufacturers] and ODMs [Original Design Manufacturers].”

    Popular development

    A very popular development with consumers would be the introduction of a dual mode handset which works with both GSM and CDMA (cdmaOne) based networks. Indeed, both Samsung and Kyocera (with the KZ850) have shown handsets offering such a capability. Both handsets are based on Qualcomm’s GSM1x technology. They will be aimed initially at the Korean market but crucially China Unicom has also announced its intention to supply dual mode handsets to its own customers — sourced from Samsung and Motorola. Clearly such handsets should soon find their way over to Europe. It’s a good example of how Qualcomm is trying to throw off its ‘CDMA only’ image as a chipset supplier. It recently announced, for example, that 14 leading manufacturers in China, Europe, Japan, South Korea, Taiwan and the United States are busy developing W-CDMA/UMTS products using its chips for scheduled 3G network launches in 2004.
    In terms of new developments in mobile handset design, despite the great media hype, handsets which actually support Wi-Fi in addition to, or instead of, Bluetooth are extremely thin on the ground. Motorola is planning to offer a Wi-Fi enabled handset for use on iDEN networks, for example. “We haven’t heard a lot about it really but since (Wi-Fi) is currently such a cut-throat business, I’m not sure we’d want to play in that market,” commented Richard Traherne, a consultant with the wireless business unit of Cambridge Consultants. His company previously spun-off Bluetooth chip specialists CSR (Cambridge Silicon Radio). What Traherne did predict was that software defined radio could play an important part in future mobile handset designs. “To date, it’s not been massively popular [with handset manufacturers] since it has the disadvantage of an overhead in terms of requiring extra resources,” Traherne said. “Although take-up is presently uncertain it can pay dividends for manufacturers since it would allow them to produce more handset variants per year from the same core engine design.”

    ‘Guts’ of a handset

    A similar approach has been taken by TTPCom with its Cellular Baseband Engine (CBE). The CBE is basically the entire ‘guts’ of a GSM/GPRS handset which runs on a single DSP. Other applications — like running an MPEG4 video clip; downloading an MP3 sound file; or running email client software – can then be run on a separate processor. Pascal Herczog, chief system architect with TTPCom, claims that separating the handset functions in this manner means that handset manufacturers won’t need to re-test a handset for type approval every time they make a slight modification to a handset’s design. Crucially this enables a handset manufacturer to produce a customised handset for one particular network operator, such as Vodafone Live!,  from a single, proven handset design. Better still, StarCore with its SC1400 chip, can physically put a DSP and an ARM processor onto a single piece of silicon — achieving a significant reduction in die size, costs and power consumption.

    While some observers might argue that chip development only relates to a small sector of the handset market, the latest figures for shipments in Europe for 2003 from Canalys prove otherwise. Between them, Nokia, Sony Ericsson, Motorola and Orange managed to ship two million units. So it would appear that chipset developments are bound to become more — not less — important in the coming year.

    Previous articleGetting the message
    Next articleDevil in the detail