Original URL: http://www.theregister.co.uk/2006/11/24/qualcomm_snapdragon_launch/

Qualcomm plays to its strengths with Snapdragon launch

And Motorola deal

By Faultline

Posted in Mobile, 24th November 2006 13:54 GMT

Comment On a financial analyst tour in the UK last week, Qualcomm CEO Paul Jacobs lamented that, on a day when the chipmaker made a string of major announcements, the commentators were only interested in asking about the licensing disputes with Nokia.

Important as that relationship is to Qualcomm's future performance, analysts would be well advised to listen rather more closely when the CDMA giant talks about new products.

In the end, what will cushion Qualcomm even if anti-trust and licensing arguments go against it will be its technological excellence, and its ability still to rely on turning out good chips that customers need to buy.

As such, last week's launch of Snapdragon, which brings advanced connectivity to consumer devices, should be given as much weight as the latest Nokia spat - while the latter may highlight Qualcomm's key vulnerabilities, the former plays to all its main strengths, such as its leading ability to integrate and shrink technology and its shrewd understanding of where operators' business models are heading, and how silicon needs to support that.

Jacobs insisted last week that he saw Intel, not Texas Instruments (TI), as the key challenge to Qualcomm, regarding the latter as too driven by the Nokia business to lead market trends or to be truly broad brush.

However, Snapdragon reinforces the view - especially with Intel pulling away from its XScale smartphone architecture to concentrate on challenging the mobile chipmakers with WiMAX and Wi-Fi - that the new cellphone generation is a battle with only two real contenders, Qualcomm and TI.

This is a generation of devices that needs to combine new extremes of compactness, low power and low cost, to support accelerating operator business cycles and cope with unprecedented price pressures on handsets, with integration of the full set of multimedia functions, including ultiple radios, television and easy coexistence with other broadband platforms such as set-top boxes.

The two majors have both dramatically honed their skills at these two extremes in the past two years, resulting in architectures like TI's eCosto and LoCosto, which combine its advanced multimedia application processor platform OMAP with VoIP and video functionality in an "out of the box" reference platform, offering high function multimedia at very low cost, addressing the consumer demands of emerging markets like China.

For Qualcomm's part, it has moved rapidly to push HSxPA, CDMA EV-DO, Wi-Fi, mobile television and various VoIP and video functions into single-chip architectures, reference designs and highly integrated chipsets targeted at advanced consumers in emerging markets.

It has taken key differentiators from CDMA, notably extreme integration and strong multimedia software like Brew, into the UMTS/HSxPA world and could also, political objections aside, exploit the same advantages in WiMAX, a step Jacobs hinted last summer would happen once Qualcomm saw sufficient market mass for 802.16 chipsets.

Where Qualcomm has generally been perceived to lag, TI has been in the application of consistent architectures beyond the handset, so tapping into the fixed/mobile convergence market and the increasing connectivity between cellphones, PCs and set-tops.

Snapdragon is an important move to build on the existing strengths and also to steal a march on TI, more naturally suited, given its set-top box business, to convergence, in this critical area of operator interest.

The Snapdragon platform is based around a 1GHz customised ARM core designed by Qualcomm, called Scorpion, paired with a 128-bit SIMD (single instruction, multiple data) processor and a 600MHz digital signal processor to accelerate multimedia applications. On the RF chip, Snapdragon will incorporate options for EV-DO, W-CDMA, HSDPA and HSDUPA, broadcast television and multimedia, Wi-Fi and Bluetooth.

Significantly, Samsung will be the first customer. Together with an announcement of Motorola as a customer for its UMTS products, made on the same day, Qualcomm managed to demonstrate how, whatever the legal and commercial battles going on with the handset community, only Nokia has the political need and desire to try to bypass Qualcomm altogether.

For the rest, their first priority is to acquire chipsets that will support the demands that operators are making of next generation handsets, notably converged multimedia at low cost and time to market, and they will buy that from any chipmaker that can deliver. Currently, that largely means Qualcomm and TI in the 3G and 3G/multimode sectors.

Snapdragon-based chipsets are scheduled to begin sampling in the third quarter of 2007 in gaming devices and PDAs as well as smartphones, as Qualcomm looks to use its integration skills to support the demand for single consumer devices that perform multiple functions.

One likely candidate to be the first Snapdragon device on the mass market is Samsung's Q1 ultra-mobile PC. If this materialises, it will be a serious blow to Jacobs' bugbear of choice, Intel, which has an advanced roadmap for ultra-mobile PCs or handtops that keeps its x86 architecture firmly in the driving seat, even as the shape and function of the PC changes out of all recognition into a truly mobile multimedia device.

"Whether handheld devices rely on an x86-based processor or another kind of processor matters less as we go forward," said Qualcomm's Luis Pineda, head of product marketing and management. "What customers really want is the ability to run their existing applications. Whether it's on x86 or RISC doesn't matter so much."

Another significant boost for Qualcomm comes with the announcement that the company will provide Motorola with UMTS chips and work with it on future 3G phones. This is the kind of endorsement that Qualcomm badly needs. Its intellectual property revenues from W-CDMA have flourished and allowed it to cross the bridge into the formerly closed community of GSM operators, but it is well behind TI in terms of tier one customers, and for all the focus on its patents business, its lifeblood remains actual shipments of silicon.

The deal is clearly a blow for Freescale, Motorola's former chip unit which has now been taken into private ownership by a consortium led by the Blackstone Group.

A period after the spin-off when Freescale was still guaranteed certain Motorola contracts has now expired and it will not be comforting to the company's new owners that its former parent has defected so quickly on the UMTS front.

Such advances are what the chipmaker needs to cushion it from the legal and licensing blows that are likely to haunt it in the coming few years, and that will allow Qualcomm to flourish even if its darkest scenarios come to pass - rapid decline in CDMA, dramatically reduced CDMA licensing charges, even a break-up of the company.

All these are worst cases, and the reality should be a far more mixed picture, but the fact is, Qualcomm is building up its arsenal for a new and more unfriendly world, and showing skill at identifying and tapping into key carrier trends. For that reason, Snapdragon and its successors will have even more long term significance than the outcome of the ongoing Nokia debates.

Copyright © 2006, Faultline

Faultline is published by Rethink Research, a London-based publishing and consulting firm. This weekly newsletter is an assessment of the impact of the week's events in the world of digital media. Faultline is where media meets technology. Subscription details here.