Your browser is not supported. Please update it.

20 February 2006

Nokia-Qualcomm battle heats up with Sanyo venture

In a show dedicated to the GSM technology family, one of the most dramatic announcements was about CDMA ‘ the decision by Nokia to form a joint venture with Japanese rival Sanyo to build CDMA handsets. This reflects the growing trend for phonemakers to join forces in order to achieve economies of scale in markets where they do not have dominant market share ‘ Sony Ericsson, for instance ‘ and like former handset ventures such as Alcatel’s with China’s TCL, it may lead eventually to a buyout by the Asian partner. But more importantly is the impact for CDMA chip supremo Qualcomm ‘ while Nokia’s determination not to quit this sector reflects the continuing strength of CDMA platforms, the new venture will almost certainly be a blow to Qualcomm’s revenues, reducing licensing fees and giving the company’s only major competitor in CDMA chips, Texas Instruments, a credible customer at last. Though Nokia hedged round the issue of who would supply the venture with chips ‘ Sanyo is a Qualcomm customer ‘ we believe that, at least in the medium term, this move is almost certainly designed to reinforce the Finnish giant’s other recent aggressive actions to try to limit the power of Qualcomm.

Nokia has long been a major thorn in Qualcomm’s side. The Finnish company has struggled to compete in CDMA handsets, partly because it has consistently refused to purchase its chips from the US player. Instead, it has aimed to use its market weight to set up a viable alternative to Qualcomm as a major CDMA chipset supplier ‘ a position that many silicon vendors have sought, including Philips, but where all have failed because of Qualcomm’s iron control of its market, its high patent licensing fees and the technical superiority its CDMA intellectual property gives it. Nokia created a joint venture with Texas Instruments and STMicro in 2003 to try to form an viable second source to Qualcomm, and although the company said at 3GSM that it would use whichever chips made economic sense, we do not believe Nokia would have agreed to the Sanyo deal without ensuring that TI/ST would be the silicon provider.

After all, Nokia has been one of the prime activists in the complaint made by a group of wireless heavyweights to the European Commission, claiming Qualcomm is violating antitrust laws and keeping competitors from entering the 3G chip market in Europe through its business practises (reflecting similar complaints made by Broadcom in the US). The complaint (see Wireless Watch October 31 2005) accuses Qualcomm of refusing to license essential patents to potential chipset competitors on fair terms, while offering lower royalty rates to handset customers that agree to exclusively purchase chipsets from Qualcomm. The companies – which also include Broadcom, TI, Ericsson, NEC and Panasonic – also charge Qualcomm with charging excessive and disproportionate royalties for essential W-CDMA patents.

Nokia’s attack, then, is threefold – promoting a competitive supplier to Qualcomm; attacking the level of the US player’s patent revenues in the courts; and probably depriving it of revenue by winning Sanyo over to TI. This is about more than just increasing handset sales, it is part of a political war that aims to reduce Nokia’s overall licensing payments in the longer term, by weakening Qualcomm’s position and forcing it to compromise.

It will be a tough battle, and Qualcomm still holds many of the cards. Though it is almost certain to have to adapt its customary licensing practises in the more competitive W-CDMA world, and in its new markets such as dual-mode Wi-Fi and OFDM platforms, it retains a stranglehold on its native sector. Some of TI’s stumbles have been blamed on the fact that Qualcomm abandoned the planned EV-DV evolution of CDMA2000, into which TI had ploughed much of its R&D, on the basis of lack of operator demand ‘ but possibly also to wrongfoot the new challenger.

By contrast with its 38% share of GSM handsets worldwide, Nokia’s CDMA share is only about 13%, though this has risen from about 5% in 2002 on the back of aggressive marketing and development. However, the business runs at a loss, which is indicative not only of Nokia’s failure to achieve the necessary scale to procure components cheaply and increase margins ‘ something that joining with Sanyo should help address ‘ but also the perils of refusing to work with Qualcomm. Since the latter owns so much CDMA intellectual property, its challengers are always struggling to catch up technologically, and incur vast R&D costs in the process ‘ out of proportion, in Nokia’s case, to the sales of fewer than 20m CDMA units a year it gains. Nokia, for instance, does not yet have an EV-DO 2.5G handset, while larger CDMA players like LG, working closely with Qualcomm, launched them last year. Nokia and TI had hoped to steal a march by focusing their R&D on the planned next generation of the CDMA platform, EV-DV, until this was effectively cancelled by Qualcomm, wasting their efforts and cash.

Given the challenges outlined above, the joint venture plan looks positive. Nokia will transfer its entire CDMA business unit, including about 250 employees in San Diego, California to the joint venture. Sanyo has about 7% share of the CDMA handset sector. “We think by putting the two together, we can get better than one-plus-one equals two,” Nokia CFO Rick Simonson commented. Even without growth, the combined companies will have 20% share, making them third to Samsung and LG, and leapfrogging Motorola’s 10%. It will also generate cost savings and economies of scale to improve margins ‘ the venture is targeting $256m profit within two year on 50m unit sales. Although the removal of CDMA will take $1.5bn off Nokia’s top line, its overall ‘ and closely watched ‘ margin will improve. Nokia said in 2003 that it required a 15% share to make the CDMA business viable long term. The partners also have complementary strengths ‘ Sanyo’s major customers are Sprint and KDDI, while Nokia has fared best in India and Brazil, while Sanyo has focused on high end models, unlike its partner, and so will bring an EV-DO phone to the fold. Financial details and brand identity have not yet been disclosed but the latter is likely to reflect both companies ‘ while possibly giving Nokia an option to exit the venture should CDMA growth not live up to expectations.

But the climate may improve, for the industry is ganging up on Qualcomm, even its traditional friends. In December, Samsung said it would launch a CDMA EV-DO handset based on its own chipset in the first half of 2006. Although it has not offered this chip on the open market, and has only sold 200,000 CDMA handsets using its own silicon to date, the architecture clearly has a significance beyond short term sales. A Samsung official told Korea Times: ‘When it comes to best selling models, the sales quantity usually passes 1m. However, the sales of these non-Qualcomm handsets have special meaning to us. They indicate that our own chipset has the same market potential as Qualcomm’s.’

However, even if Qualcomm is forced to cut back on its licensing fees, it still holds a huge store of fundamental intellectual property and nobody can rob it of the technical headstart that entails, or the company’s genuine engineering brilliance. This will enable it to win customers in W-CDMA even among phonemakers that are politically hostile, but cannot afford to settle for an alternative solution, for the sake of a principle, if it offers lower performance or higher cost. For instance, a new research note from Caris & Company indicates that Qualcomm is likely to become the chip supplier for the latest generation of Motorola’s HSDPA phones because Freescale, the spin-off from Motorola, has not come up with a sufficiently small chip for the ultrathin handsets.

“Freescale is having difficulty reducing the size of its chipset. Qualcomm could become the supplier for Motorola’s 3G phone,” said Caris.

Unlike many prominent patent owners, Qualcomm can compete effectively in terms of products delivered as well as licensing strangleholds. The tactics of Nokia and TI are only a taster for the real battle royal that will rage over the more open battlefield of W-CDMA and HSDPA.

BOX ‘ STMicroelectronics

Another beneficiary of Nokia’s huge purchasing power is STMicroelectronics, but the chipmaker is looking to reduce its dependence on the Finnish giant and to focus more strongly on 3G.

The company, which retreated last year from a major effort to develop a GSM baseband chipset, said its growth is now coming from 3G penetration and the mixed signal business and that it is now the leading producer of 3G RF and energy management products. ST is also anticipating big growth for its 3G digital baseband technology, which will start volume production by the end of 2006.

“Much of its engineering resources came from a team which stopped 2G chipset development last year, and a key asset we acquired from Alcatel,” said the company.

While Nokia remains the chipmaker’s biggest customer, expanding its customer list has been high on the agenda, and it now has four customers, including Nokia, engaged for its second generation Nomardik application processor, STn8115, and has won a second client, the Ericsson Mobile Platform/Philips Semiconductor chipset, for its camera modules. Bluetooth and Wi-Fi chips are also driving ST’s mobile platform with Nokia, Sharp and Sagem signed up.