The Broadcom Qualcomm war is far from over. What has happened this week is that it has been taken out of the courtroom and placed firmly back into the marketplace.
We admire both Qualcomm and Broadcom, which, given we don’t admire many companies at all, is quite a feat. Qualcomm is innovative in the extreme, creates new markets and looks at problems a new way, and applies circuit design rigour to solving real world problems. It also attacks with aggressive marketing. Broadcom on the other hand does not create markets, but harvests them. It has the ingenuity to commoditize any chip market and is the strongest survivor of the Wi-Fi price wars. It has applied the same rules to set tops, DVD player chips, DSL and cable modem circuitry as well as Bluetooth. It will also attack things like ATSC broadcast receiver chips in the near future. Each time it enters a market it comes away with a significant market share.
Our simplified view of the four year squabble, settled this week between the two, dragged through US courtrooms, the International Trade Commission and a variety of overseas anti-trust authorities, is that Qualcomm saw Broadcom entering its market and said ‘We will stop this.’
One of the tactics which probably created this dispute is that Qualcomm implied to existing CDMA chip customers that using Broadcom chips would lead to bigger royalty charges from Qualcomm. Even saying that in print will lose us the few friends that we have inside Qualcomm.
But if we lay aside tactics, legal or otherwise, for a second, we have to reach the conclusion that Broadcom won this dispute, since it now clearly has a license to ‘existing’ Qualcomm technologies, but without seeing the paperwork (and no-one outside the two will) we cannot know if this includes future Qualcomm created technologies or not, and if so, how far into the future.
But for now CDMA is NOT a growing market, but a falling one. It is still large and still 90% dominated by Qualcomm, quite simply because other silicon companies have tried to create CDMA chips and found that they either can’t make them work or that the Qualcomm chips are either a) better b) cheaper or c) Qualcomm charged the customer more for royalties and so put them off or d) all of the above.
This is not the case in UMTS 3G chips, and the bigger effort for both companies is what they do about UMTS, a form of wideband CDMA, very different from the Qualcomm invented CDMA.
Qualcomm, for its part, is getting more and more of its revenues from UMTS, specifically due to its ability to get to every sub-market first, such as higher speed HSDPA data delivery chipsets which will go from the 3.6 Mbps widely available today, through the 7.2 Mbps level that is going through deployment now, to the 14.4 Mbps and 28 Mbps and 42 Mbps which are all on the 3GPP roadmap. By getting there first in each chip generation Qualcomm is taking market share from all the other chip builders. This is the market that Broadcom really needs to attack eventually but in the meantime it also needs to attack older GSM markets and commoditise these chips, for older networks. It will take some years before GSM is overtaken in chips required per year as a market, even if the value of supplying chips to 3G markets is already higher.
While that war unravels, and let’s face it, it will be fought in the R&D labs of the two companies, as well as among new Chinese upstarts, the Status Quo has been retained by Qualcomm.
This means that the deal with Broadcom, like the one with Nokia before it, has been structured in such a way that if another of Qualcomm’s customers ‘want the same terms’ it will not mean that Qualcomm’s royalty payments will fall dramatically. That’s the BIG thing at Qualcomm, where far too much of its profit comes from a general royalty deal with component makers and device and equipment makers. In order to retain that as a model, it has had to cut deals with Nokia and Broadcom that do not undermine its general royalty terms, especially to the larger customers that have demanded most favoured nation clauses.
So that’s a victory to Qualcomm that no court would have given it. It could never had settled inside a court, because the terms would have been very public and not reached by negotiation. This is why we always knew that it would settle with Broadcom. Any other kind of settlement might have created a serious challenge for over 50% of the company’s profits ‘ so a very short term and immediate problem.
However, with that out of the way, Qualcomm can get on with business. It can create new markets and its chip design excellence will still mean that it can dominate markets that it doesn’t create ‘ like UMTS and of course its next iteration LTE.
But it also means that Broadcom cannot be stopped at what it does ‘ create a commodity from a rarity and drive prices down all over the cellular market.
The $891 million payments over four years that this settlement gets Broadcom from Qualcomm is really only a short term gain for Broadcom. It means that it has a moral victory and a little cash. But the long term victory for Broadcom is that it has a place at the cellular chip table with impunity, and no-one looking over its shoulder. It was probably a much harder market for it to enter than it expected, but now it is in, it can potentially double its market share in the next year or so, as long as it can make the chips in volume and to a high enough yield and standard.
But there remains a number of issues. Once an anti-trust door is open, the withdrawal of a complaint is the last reason why it should be closed. Microsoft found this to its displeasure when it paid off RealNetworks and Sun Microsystems and the European Commission proceeded with the complaints against it. The same will happen to Qualcomm. Americans always think that this is the same as a civil suit and once the complainant is happy it is done and dusted. But that’s not how the Commission sees things. If the Qualcomm terms were iniquitous to Broadcom, they could also inhibit other rivals and therefore slow the market, to the detriment of the consumer. So there is every chance that eventually a Qualcomm ruling on anti-trust and double dipping royalties will be made by the European Commission. Qualcomm cannot do without Europe because it is over a third of the world’s cellular chips, so if that happens, it will have to comply. The same can be said of the Korean anti-trust complaint. Although Europe speaks for over 500 million phone owners, Korea perhaps 40 million.
So while Broadcom cannot continue with any complaints against Qualcomm and while the biggest handset maker in the world in Nokia is happy with Qualcomm now, that it NOT the end of it. However these things take time and it probably will not re-emerge for 18 months or more and may require fairly innocuous changes (or it may not) to Qualcomm business practises.
For now the deal is that there is a settlement for the multi-year patent agreement which dismisses all litigation between the two companies.
They have each granted each other ‘certain’ unspecified patent rights and Qualcomm will pay Broadcom $891 million over a four-year period.
The deal will not change Qualcomm’s licensing revenue model. The US import blocks on Qualcomm chips will now evaporate which caused Sprint and Verizon some issues on adopting new handsets. The deal is restricted to the cellular space, no Bluetooth and other technologies are out of the equation. Broadcom cellular chip customers still have to pay Qualcomm royalties in the same manner as other licensees.
Interestingly Broadcom stock fell on the announcement while Qualcomm leapt on Monday and then fell back. Both stocks are probably now in the grip of fears over semiconductors and this deal is now in the price.
Almost as if Qualcomm was underlining its point about its royalty model remaining intact, it issued a statement today about signing a new royalty deal over CDMA and TD-SCDMA (the new Chinese standard) with Hong Kong phone maker Sanmu Communications, presumably mostly for the bulk Chinese market.