Qualcomm continues to rebuff Broadcom’s advances, but is set for a Valentine’s Day meeting, the first time it will have agreed even to discuss the fellow chipmaker’s takeover bid, which was raised last week to £121bn.
The prospects for a deal which would significantly reshape the mobile chip industry have looked poor so far, with the two companies engaging in wars of words and aggressive moves to persuade Qualcomm shareholders of their case. However, the increased price, and Qualcomm’s continuing difficulties with Apple and other legal challenges, may tip the balance in favor of Broadcom.
That outcome would almost certainly lead to settlement of the litigation with Apple and other clients and antitrust bodies, but at the expense of Qualcomm’s famous licensing model. Broadcom has already hinted it would run that aspect of the business very differently – between the lines, in a way that would appease the organizations which are currently attacking Qualcomm’s business practices. But while the licensing approach has often been unpopular, there is a very strong risk of throwing the baby out with the bathwater – in sacrificing the contentious practices, Qualcomm’s fabled engineering and R&D activities are also likely to be scaled back, and the industry would then lose a major source of innovation and creativity. That, in turn, would cede even more ground in mobile innovation to China and its up-and-coming chip designers.
Initially, the increased bid price did not seem to effect a thaw in relations (see Wireless Watch February 6 2018). The heads of both companies went on the offensive, Hock Tan of Broadcom saying Qualcomm’s business model needed to be “reset” to reflect changes in the industry; Qualcomm’s chairman Paul Jacobs retorting that the offer ascribed “no value for the significant opportunity in 5G.”
“Certainly it’s an offer that provides more value to the shareholders of Qualcomm than any standalone value that Qualcomm has tried or may think they can try to create,” Tan said in an interview with the Wall Street Journal. “We are asking the board of Qualcomm, and we believe any rational board would engage with us. And if they don’t, we want the Qualcomm shareholders to have the ability to elect our responsible slate of nominees who will then act in their best interests.”
He said in the interview: “We have to reset the business model. We would not have announced the transaction if we were not very confident that customers would embrace the proposed combination.”
He denied claims that Broadcom would reduce the innovation and R&D spend for which Qualcomm is so well known, saying: “On the issue of innovation, that’s the second level of FUD that’s been thrown. We probably out-innovate any competitors out there in the product lines that are core to Broadcom because we’re a technology-based company.
We focus on our core business and sort of stick to our knitting.”
And he was certainly not pulling any punches in discussing the licensing business. He accused Qualcomm of being too heavily driven by engineers. He said: “The trouble is the people who run it. They’re still fighting yesterday’s battles. Their revenue is virtually flat. The world has changed. The engineers who run this business have not changed.”
He went on: “You may start off as an IP licensing business, at the end of the day an IP licensing business is very hard to sustain. It’s hard to be a patent troll forever. You create more value for a customer by translating, embedding this technology you develop into silicon, hence semiconductor.” Of course, one accusation that is never seriously made about Qualcomm’s controversial business structure is that it is a troll, given that so much of its R&D is turned into products, but Tan clearly wants to appeal to shareholders’ and customers’ understandable nervousness about where the litigation may land Qualcomm.
He added: “You can’t run a business this way where you’re constantly fighting with regulators and your practices are viewed as anti-competitive and two of your major licensing customers have decided they don’t want to pay you the money and are taking litigation. That’s not a sustainable model … Nobody fights with their biggest customer.”
Jacobs was more measured in his tone, but made his views clear in an open letter to Tan. He wrote: “Your proposal is inferior relative to our prospects as an independent company and is significantly below both trading and transaction multiples in our sector. Your proposal ascribes no value to our accretive NXP acquisition, no value for the expected resolution of our current licensing disputes and no value for the significant opportunity in 5G.”
However, despite the fine words, Qualcomm has now agreed to a meeting. “The Board is committed to exploring all options for maximizing shareholder value, and so we would be prepared to meet with you to allow you to explain how you would attempt to bridge these gaps in both value and deal certainty and to better understand the significant issues that remain unaddressed in your proposal,” Jacobs wrote to Broadcom.
Tan still couldn’t resist another jibe, answering: “Following Qualcomm’s announcement today that it is willing to meet with us, we offered to meet with Qualcomm on Friday, Saturday or Sunday. I was astonished to hear that Qualcomm is not willing to meet until Tuesday.”
Regardless of the politics and the posturing, Qualcomm’s challenge now revolves around what Jacobs described in his open letter – instilling confidence in shareholders that, despite its problems, the company can deliver more value as an independent, particularly because of its proposed acquisition of NXP, which will give it a far bigger position in the automotive and Internet of Things sectors; and its 5G leadership.
It was timely, then, that in the same week as the exchange of letters, Qualcomm was announcing its biggest commercial breakthrough in 5G, with 19 vendors officially saying they would use its Snapdragon X50 5G modem, and 18 operators agreeing to conduct 5G field trials with it. However, Apple and Huawei are not on the OEM list, even though they are both current customers, and they account for a sizeable percentage of the smartphone market.
Huawei’s HiSilicon chip division is expected to design a 5G modem of its own, possibly working with its strategic partner, MediaTek. It remains to be seen whether that will satisfy all the requirements of the Chinese firm’s increasingly successful handset business – now third in the world – or whether, like Samsung, it divides its favors between Qualcomm and an inhouse solution. Samsung has a 5G modem and system-on-chip of its own in the works and, if that is a successful product, it will surely try to reduce the number of models in which it uses an external product. However, it had previously announced that it was working with Qualcomm on 5G, along with a new patents agreement.
Apple is still a major Qualcomm customer too, despite the lawsuits, but would be very unlikely to put its name on a public list of endorsements while litigation is raging, and is reported in some quarters to be planning to transfer all its advanced smartphone modem business to Intel.
The uncertain future of relations with the top three smartphone makers are clearly a risk factor, though Qualcomm has signed up some of the newer Chinese players which could be jostling for top positions in 5G, including Oppo and Xiaomi. These, notably, were also on a recent list of vendors which pledged support for Qualcomm’s newest RF front end solution, suggesting that these companies are happy to fill as many slots as possible from one supplier if that gives them a fully integrated solution which is, therefore, easy to deploy, reducing time to market and power consumption.
And some of the OEMs on the 5G modem roll call point to Qualcomm’s hopes to offset the slowdown in smartphone growth, and the challenges at the largest vendors, by pushing Snapdragon into all kinds of other connected devices with higher growth potential. For instance, some of the names are NetComm Wireless, Netgear, Sierra Wireless, Telit, Wingtech and WNC. Many of these are known for home WiFi equipment – a strong Broadcom business – and other relatively low volume items (compared to handsets), but almost without exception, they are expanding their businesses into modules and devices for the IoT, which could lead to far greater numbers in future (though the pricing and margins for many of these products remains highly uncertain at scale).
Other 5G customers include LG, ZTE, Sony and HTC from among the traditional handset names, as well as Asus, Fujitsu, HMD Global (which uses the Nokia brand) and Sharp.
Qualcomm also said that 18 carriers have committed to use the X50 chip in sub-6 GHz and millimeter wave field trials, some of them planning to kick off commercial 5G services next year, when the chip giant expects the first mainstream smartphones to be available. The operators include AT&T, BT, the three Chinese carriers, Deutsche Telekom, KDDI, KT, LG Uplus, NTT Docomo, Orange, Singtel, SK Telecom, Sprint, Telstra, TIM, Verizon and Vodafone Group.
The extensive number of Chinese names on both the OEM and operator lists highlights that Qualcomm’s relationships in that country have improved greatly, and the payments it made in 2015 to end a regulatory probe – during which many licensees withheld patent payments – were money well spent. The firm says it expects to be only weeks away from Chinese approval for its purchase of NXP, the last regulatory hurdle to a deal which could strengthen the confidence of Qualcomm’s investors to resist a Broadcom takeover, by making the company number one in automotive overnight, and strengthening its hand in the new IoT markets.
The X50 is made up of a baseband chip and a transceiver module, the latter incorporating an RF front end module with an envelope tracker, power amplifiers, and integrated antennas. The solution demonstrates something at which Qualcomm has always excelled, integrating complex components tightly into a compact, low power unit. In the X50’s case, the elements are packaged in a single board that “looks like a chip and a module but it’s really a couple of chips and an RF front end module with antennas,” as Peter Carson, senior product management director, put it. Qualcomm expects to reduce the size of some elements, notably the millimeter wave module, by at least 50% for commercial products.