Qualcomm Inc (QCOM)

[khemara_qc]

Qualcomm's bold move
http://p2pnet.net/story/2943

Caroline Gabriel ----
---- Wireless Watch
p2pnet.net News Feature:- Qualcomm has always been a company of extreme audacity. Having taken on the prevailing GSM world to establish its own de facto standard in CDMA, it has created a business around its vast store of intellectual property that, if its new results are anything to go by, has far more life left in it than naysayers would have predicted even a year ago. CDMA networks are proving resilient in face of the long and complex roll-out of UMTS; Qualcomm still has significant patents involved in that standard, and is seeking a UMTS power base of its own; and now it is leaping ahead of all wireless contenders into the nascent market for the triple play.


The chipmaker has to be bold. Its dependence on CDMA means the eventual move to all-IP networks, and the likely rise of OFDM and WiMAX, are major threats, and explain its implacable hostility to 802.11 and 802.16. Therefore it needs a game plan that will enable it to shift to new revenue streams without sacrificing its core expertise and partnerships.


The boldest part of this plan to be revealed so far is Qualcomm’s creation of a national US broadcasting network for use by carriers and other partners, which will boost uptake of CDMA and W-CDMA cellphones and, potentially, other devices that Qualcomm will devise over the coming few years.


Qualcomm plans to invest $800m over four to five years in a new subsidiary that will create a network to help carriers bring television to smartphones. While other chipmakers such as Intel mull whether to invest in spectrum in order to further their wireless strategies, Qualcomm has gone ahead and done it, buying up licenses to broadcast in the former analog television bands in the US. This mirrors its aggressiveness in buying up, through an affiliate, 450MHz spectrum in Europe and other areas to bolster the chances of CDMA in those bands.


Qualcomm will initially target 30 US cities for its MediaFLO USA network, named after its content distribution system of the same name, announced in March. Interestingly, Qualcomm also announced last month a video multicasting system for MediaFLO, simply called FLO (Forward Link Only), which uses OFDM as an alternative to the usual CDMA base. This shows the chipmaker, despite all its powerful reasons to lock its gates against any non-CDMA technology, recognizing that, for some applications, OFDM may be superior, particularly for delivering more channels of content at lower cost. While CDMA is more spectrally efficient, in general, than GSM, it cannot deliver the bandwidth of a WiMAX. If Qualcomm continues to demonstrate willingness to open up to outside technologies, it could stand an improved chance of winning in the multi-network, multi-protocol world of 4G, rather than seeking to defend an embattled CDMA kingdom against impossible odds.


In such a multi-network world, its royalty revenues would fall, but by launching added value services such as MediaFLO at an early stage, it could position itself for a far wider influence.


MediaFlo is based on OFDM outbound high speed technology and provides a one-way, high speed system to deliver audio and video. Multiple channels within Channel 55 will be made available to both CDMA and W-CDMA carriers as well as MVNOs and content resellers. This could leave the CDMA or W-CDMA channels free to deliver higher quantities of two-way data. TV broadcasters could license the network to expand their viewing base to mobile users, particularly those, like sports channel ESPN, that own content as well as distribution. Within its 6MHz spectrum, MediaFlo will be able to deliver 50-100 national or local content channels. The MediaFLO system will be an end-to-end solution, from the content distribution system through program guides and billing to the 700MHz channels.


Like Intel, Qualcomm is not seeking to be a network operator but to use spectrum to boost uptake of, and new applications for, its core revenue earner (hence the plan to spin off the unit once it has achieved its goal). Just as Brew aggregates download content for operators, handling distribution, billing and content provider relationships, so MediaFlo – which may eventually be spun off to investors – will act as an aggregator of television content for carriers, striking deals with TV stations and other parties. The network will start operating in 2006 and the first carrier to support it is Sprint, already a firm believer in mobile TV as a future revenue driver in the US.


Paul Jacobs, head of the unit that houses MediaFLO, said the FLO OFDM-based technology uses high broadcast towers and high wattage to cover a city with an average of 2-3 transmitters, using the former analog TV spectrum in 700MHz, which should be vacated by 2007-8 as the broadcasters shift into digital TV.


Qualcomm has bought the nationwide license for one of the UHF-TV channels becoming available, Channel 55, but other parties, notably the WiMAX community, have their eyes firmly on other channels in 700MHz, which lends itself to mobile broadband wireless services. It will be ironic if, just as the old broadcasters move out of their former spectrum, it is colonized not by conventional data services, as originally envisaged, but by OFDM-based networks aiming to deliver the full triple play, including television. Jacobs, like most of his peers, does not expect consumers to watch full TV programs on their phones, but envisages a new breed of more viewer-friendly wireless devices as well as growing popularity of short videos and clips. But longer term, if Qualcomm is not entirely wedded to CDMA, why the carriers leasing its MediaFLO network not seek – regulators willing – to deliver their video and TV content into consumers’ homes too, via set-top boxes and other devices? That is certainly the plan for WiMAX – the question is whether FLO will support the economics, in terms of subscriber unit design, and will attract the content partners to make it attractive to triple play wireless operators.


All this is part of Qualcomm’s bid to build on its success in CDMA and colonize new wireless markets. Qualcomm wants no less than to be a “wireless Intel” – just as Intel itself turns increasingly to wireless. While Intel, despite its cellphone architecture, sees its real wireless and mobile play as revolving around WiMAX, IP and OFDM, Qualcomm seeks to expand on its domination of the 30% of the cellphone sector that is based on CDMA by moving aggressively into the other 70%, which is moving from GSM to W-CDMA. Qualcomm, despite its very small showing in GSM-based markets to date, claims its goal is a 50% stake in W-CDMA chips, which would mean overtaking the mighty Texas Instruments.


While these ambitions seem overblown, Qualcomm certainly has the intellectual property, outsourced manufacturing partners, channels and sheer temerity to make a strong headway in W-CDMA. It will never have the clear run – and therefore the high price margins – that it enjoyed in its home base, but it needs to expand quickly, now that it faces its first ever viable challenge in CDMA from TI itself.


Merrill Lynch has stated that, if Qualcomm achieves 30% share of W-CDMA by 2007, as that market grows from 14m to 204m chips a year, earnings per share could easily swell from $1.20 in fiscal 2005 to $1.75 in 2007 - 46% growth over two years. Even 30% seems over optimistic, given TI’s strength and other challengers such as Freescale looming. But Qualcomm has the additional boost of its patent revenue – it owns about 35% of the patents underpinning W-CDMA, and the technology already accounts for 25% of Qualcomm’s total royalty revenue, indicating the high level of prices that the chipmaker is able to command for its intellectual property.


Handset maker support will be critical, since none of them are required to turn to Qualcomm as they are in CDMA. In that camp, only Nokia was strong enough to resist, making its own CDMA chips until it enlisted TI to its cause as a second source. In W-CDMA, Qualcomm has so far signed up Siemens, a confidence booster that should draw the attentions of others, notably Samsung, which is already close to Qualcomm in CDMA – and was, indeed, the critical supporter in making that technology a success in its first power base, South Korea. Also important is Qualcomm’s W-CDMA joint handset development program with Japan’s NTT DoCoMo, always an early mover in new mobile services and the largest W-CDMA carrier.


But Qualcomm, like arch-rival TI, knows that pure chipmaking is no longer a business that will deliver the type of growth and margins it craves. Instead, both giants are developing their multimedia capabilities aggressively – TI with its OMAP2 architecture and projects such as its mobile phone television chips.


Similarly, Qualcomm is transforming itself into a mobile multimedia company, though with a stronger software edge than TI. The television venture is the most audacious step, but there are many smaller signs, such as the company’s recent acquisition of Trigenix to boost its user interface software. These capabilities will be built into the Brew content download and management platform, which is taking a central position in Qualcomm’s long term strategy that explains and justifies why the company has clung for so long to the product. A year ago this seemed inexplicable – Brew, though popular with many CDMA carriers, was being dwarfed by Java and was a small part of the overall Qualcomm business. Now, as it shifts into multimedia software for handsets and into providing the core of the carrier content system, the attachment to Brew makes perfect sense.


Qualcomm is working on chips for every aspect of multimedia usage. For instance, it is using ATI’s 3D graphics chips to support mobile gaming at a new level by 2006, and is developing a chip, also slated for 2006, to beam photos directly from a handset to a nearby television. The company even aims to incorporate digital rights management into the chip to allow users to download movies to a handset and watch them by pointing the device at a television.


Such moves show that Qualcomm will not lie down and take a decline in the CDMA market passively, even as TI starts to bite into its strongest sectors. But, despite its current strong financial showing, it faces many short term hurdles before it takes on the largest risk of its television venture and W-CDMA push. Near term issues, which drove the chipmaker to warn, at its recent results announcement, of a likely “slow start to 2005”, include a slowdown in two of its most important CDMA territories, China and India. Qualcomm’s outsourced-only manufacturing model carries supply risks and it currently faces shortages of some key chips. Also, W-CDMA royalty revenues are harder to predict than CDMA ones, and TI is eating into the core CDMA sector at last.


Whether or not Qualcomm’s ambitious venture succeeds, we can’t help thinking that the WiMAX community could learn a great deal from its greatest critic. Instead of complex politicking, Qualcomm has just gone out and done it, and has pointed the way for a model where cellular networks and OFDM can coexist, playing to the greatest strength of each and providing a new business model for operators that does not cannibalize the existing network. Should WiMAX be able to penetrate TV spectrum too, the pressure will be on to come up with equally inclusive and ingenious models with which to attract the wireless triple play operators.

Caroline Gabriel - Wireless Watch, UK

(Tuesday 9th November 2004)