Qualcomm considers radical action as pressure mounts
Chipmaker feels the heat in face of likely EU probe
Analysis Pressure is mounting on Qualcomm on all fronts, with CEO Paul Jacobs hinting for the first time that he might consider breaking up the company.
Although he made it clear this would be a last resort, the usually sanguine chipmaker is clearly feeling the heat of intensified calls for probes of its business practises (with a full EU investigation looking likely); a fresh round of legal battles with Nokia; and a new setback – an unprecedented decision by the IEEE standards body to suspend the activities of one of its taskgroups, 802.20, alleging undue influence by Qualcomm.
A radical rethink of processes and structures, especially as regards intellectual property licensing, is clearly essential for Qualcomm to answer its critics, and avoid a Microsoft-style situation where it spends years taking a defiant stand through a series of investigations that are costly to both the coffers and the reputation.
Like Microsoft, Qualcomm stands accused of tying together two aspects of its business – in this case, patent licenses and chip sales, rather than operating systems and applications – to achieve a quasi-monopoly and make it impossible for other companies to compete on a level playing field.
Arguably, unlike Microsoft, Qualcomm – whatever the rights and wrongs of the debate over its current business practises – would certainly be a success story even if its different activities were separated. It is an effective chipmaker, as the announcement of yet another improved quarterly forecast last week demonstrated, and has genuine technological advances that companies want to license, sometimes in spite of their political instincts (O2 signing up for Qualcomm's user interface platform, for instance).
So break-up need not be disastrous, though it would certainly reduce Qualcomm's advantages over its rivals as it seeks to extend its power beyond its native CDMA sector and into the whole 3G plus market – including chips for W-CDMA and, in future, China's TD-SCDMA, plus chips and patents in OFDM technologies.
Jacobs told the media last weekend that if a break-up were necessary, "then that's what we're going to do. But we expect that we're going to be able to keep the company together".
Following complaints about Qualcomm's business practises made to the EC last October by Ericsson, Nokia, Texas Instruments, Broadcom, NEC and Panasonic, the parties are waiting to see whether the Commission decides to launch a formal probe. Jacobs admitted: "I guess I won't be surprised if there is an investigation."
Broadcom has also filed anti-trust suits and complaints in the US, claiming that Qualcomm ties its intellectual property licensing and chip contracts together unlawfully. The key complaint in Europe is that Qualcomm is not reasonably licensing its patents, charging the same rates on W-CDMA as on its own CDMA2000 even though its share of the intellectual property is far smaller.
Jacobs responds that the royalty rate is a low single-digit percentage, and the company makes a virtue of a single rate that covers all its platforms, including the emerging OFDM products.
The OFDM issues
These products and their associated patents – most importantly the Flash-OFDM broadband wireless network acquired with Flarion and the FLO mobile broadcasting system – are critical to Qualcomm's bid to retain a key role in next generation mobile communications, even as the 3G platforms, all underpinned by its core CDMA technology, gradually give way to a far more varied 4G picture, with OFDM the key foundation.
Qualcomm has set about drumming up industry support, and licensing revenue, for its OFDM architectures in anticipation of this shift. On the mobile TV side, the FLO Forum is growing (see inset) and the chipmaker has taken a lead in multi-radio integrated chipsets that support not just MediaFLO but other likely standards (see separate item about Siano).
This reflects a Qualcomm that will be less religious about its own technologies and more ready to embrace open standards in order to drive new revenues, and to keep CDMA healthy by allowing it to coexist with other networks such as Wi-Fi.
On the Flash-OFDM front, however, Qualcomm has not been able to avoid a religious war. In the broadband wireless market, it is promoting a 3G-plus vision based around retaining CDMA as the premier platform for mobility, complemented by OFDM for some high bandwidth multimedia applications plus backhaul. In this, it is seeking to downplay the challenge from WiMAX, whose domination by Intel makes it anathema to Qualcomm - and one of whose objectives was to create a next generation wireless architecture with no reliance on Qualcomm patents.
Qualcomm cannot embrace WiMAX as it (belatedly) has Wi-Fi, because 802.16e aims directly to challenge the dominance of the CDMA family. The CDMA supremo can only support WiMAX if 802.16 in one of two scenarios – if WiMAX is no longer a threat, perhaps if its role remains confined to fixed and portable access, so it can be complementary to CDMA as Wi-Fi is; or if Qualcomm can gain a strong position of its own in WiMAX patents and chip sales.
While the chip giant claims intellectual property within 802.16, and has so far scored two licensees to support these claims – Soma Networks and, last week, UTC (see inset) – it is unlikely to hold a commanding patents position as it does in CDMA. So it has been seeking to create a rival to WiMAX, and a more assured licensing business, around Flash-OFDM and other OFDM patents it holds.
This centered on a bid to revive the moribund activities of the 802.20 taskgroup – whose efforts to create a mobile broadband standard, nicknamed Mobile-Fi, had been largely eclipsed by the increasingly mobile remit of 802.16e, but which had always been heavily focused on the Flarion architecture.
Though once an opponent of Mobile-Fi, Qualcomm now saw its chance to dominate an OFDM standard as Intel did WiMAX in its early days.
Setbacks in the IEEE
However, the IEEE has dealt a blow to this strategy, with a move that the 802.16e faction has been demanding all year. The standards body's SA Standards Board has temporarily suspended deliberations by the 802.20 working group after representatives from Intel and Motorola threatened to file formal complaints about the way the working group's chairman, Jerry Upton, allegedly handled draft proposals in favor of Qualcomm and Kyocera.
Steve Mills, chairman of the Standards Board, told reporters that Upton, officially listed as an independent consultant, had confirmed to the IEEE that he did have a relationship with Qualcomm (Intel says Qualcomm pays him).
Intel has also alleged that Qualcomm had brought people to meetings in order to ensure a majority (technically against IEEE rules, though in practise hardly unusual. In some past taskgroups, packing of meetings has been a spectator sport).
A Qualcomm spokesman said: "Since the IEEE operates on a one-person, one-vote basis, it is not uncommon for companies to bring people to a working group, and what has been alleged for 802.20 has been evident in many 802 organisations."
The company also said that, prior to the complaints, only two draft proposals had actually been submitted, one from Qualcomm alone and one from Qualcomm working with Kyocera.
The latest row within the IEEE highlights how standards groups are increasingly used to play out vendor politics and rivalries, and will further throw doubt over how effective the IEEE process is for the modern wireless industry – especially in the wake of the failure of the 802.15.3a UltraWideBand effort and various hitches in the 802.11n fast Wi-Fi process, all created by the politics of large chipmakers.
It is probably rather less damaging for Qualcomm itself, since the chances of passing 802.20 off as a truly open standard were always slim. It would always have been regarded as virtually a Qualcomm technology anyway, and Flash-OFDM would be better seeking industry support through a backers' club like FLO Forum which can provide a measure of openness, without being bogged down in standards body procedures.
After all, investigations by the EC, IEEE or any other body, while damaging to a vendor's reputation and potentially significant in the future, move slowly and are rife with politics and factional interests. It is far more urgent that the vendor can make its case to the customers, and maintain a sufficiently strong business that even a drastic decision by the authorities – such as a forced breakup – does not prove disastrous.
So the patents licensing debate that matters most urgently is not taking place in the corridors of the EC or Department of Justice, but with the major operators in Qualcomm's key markets.
Korea, once the country that most helped establish CDMA as an alternative to GSM, has been chafing against the level of royalties its vendors and operators pay to the west in general and Qualcomm in particular. China is claiming that it does not need to pay royalties to Qualcomm for its homegrown 3G technology, TDSCDMA, an assertion the chipmaker disputes.
And now another giant, India, has joined the battle, pressuring Qualcomm to lower its royalty rates for CDMA. The Indian government and operator Reliance Communications - the largest CDMA provider in the subcontinent and responsible for 8 per cent of the global CDMA user base - are complaining that CDMA is not affordable for India, prompting CEO Jacobs to pay a visit to the Indian communications minister, Dayanidhi Maran.
He is likely to be looking for a concession similar to the one Qualcomm made a few years ago in China, lowering the royalty rates for that country in order to stimulate CDMA uptake. Indian officials said the royalty on CDMA handsets sold in India is 7 per cent.
These discussions are vital and may indicate the first, most urgent changes to the Qualcomm approach. More flexible licensing regimes, or a reduced dependence on patents revenue at all, may be the price of ensuring that CDMA-based technologies take off as rapidly as possible in high growth markets.
Unlike responses to possible EU investigations, these decisions need to be made quickly. They will surely spark a broader review of Qualcomm's business models to prepare it for the 4G world and deflect some of the mounting pressure on the existing model that has served it so well until now.
The company has sufficient strengths and technology assets to make significant changes without destroying itself, and some clever adjustments made now - even if these include some concessions to powerful governments and even to powerful handset makers – should ensure it does not have to go as far as a break-up, and that it still stands a fighting chance of retaining a powerful market position in 4G.
Copyright © 2006, Wireless Watch
Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.