You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!

  • Not-so-ulterior motive leads SoftBank to ARM

    This week’s £24.3B offer for ARM Holdings plc from SoftBank has been widely viewed as Brexit reflexit. It did firm up in the preceding two weeks, but this acquisition offer has been years in the making – and if it sticks, one SoftBank motive many analysts and editors are missing comes front and center.

    Of course, it may not stick. Two obvious possibilities are the British government blocking the deal, and another bidder coming in. The former came off the table almost immediately: freshly-anointed UK Chancellor of the Exchequer Philip Hammond has come out strongly in favor of the SoftBank offer, even as he and new Prime Minister Theresa May have said they will stand in the way of Brexit-motivated corporate raiding as the pound sterling contracts. (BTW, I read a headline somewhere saying Hammond and May had endorsed the ARM deal, and my first thought: what about Clarkson? #NotThatHammondAndMay)

    <blockquote class="twitter-tweet" data-lang="en"><p lang="en" dir="ltr">This would be largest ever Asian investment into the UK &amp; would double size of ARM's UK workforce. Big vote of confidence in British business.</p>&mdash; Philip Hammond (@PHammondMP) <a href="https://twitter.com/PHammondMP/status/754925558065143808">July 18, 2016</a></blockquote>
    <script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>
    Despite a lot of wagging, the plummeting British currency didn’t really factor into this deal. ARM’s stock rose 17% post-referendum, but according to the Wall Street Journal if ARM were priced in yen, it has declined 15% over the last year against the stronger Japanese currency, leading some to say ARM was “cheaper”. SoftBank CEO Masayoshi Son flat out told analysts: “I did not make the investment because of Brexit.” Son also said there have been intermittent discussions over the last several years about ARM and SoftBank doing something together. The 1700 pence offer price was likely set against the ARM all-time closing high of 1205 pence in March 2015.

    Article: A Brief History of Cadence Design Systems-arm.l-5-year.jpg
    courtesy BigCharts.com

    The timing of the ARM deal is more likely due to recent SoftBank moves. They just finished selling their stake in Supercell Oy (makers of games Clash of Clans and Boom Beach) to Tencent, and unloaded a major chunk of Chinese e-commerce behemoth Alibaba stock, locking down something around $18B in gains. Analysts hoped that capital would be used to pay down debt SoftBank took on in buying 80% of Sprint in 2013. Instead, they went after ARM, with Son citing more confidence in the Sprint situation.

    It’s not the first time Son has come after a UK tech company. SoftBank acquired the Japanese operations of Vodafone in 2006 for $15B, and has been rumored on and off to be considering buying into Vodafone itself to get a foothold in Europe alongside mobile operations in Japan and the US. I’d speculate that was the reason behind the rapid endorsement of the offer for ARM from Hammond.

    The other possibility is more complicated. At first, and I even went there, it would seem the big stakeholders in the ARM ecosystem would have something to say, possibly with a competing bid. Apple certainly has enough cash lying around to make this happen, and given their long history with ARM there is precedent there. Samsung also has no shortage of capital and similar ties. Qualcomm would seem to have interest but would be taking on a lot of debt at a bad time. Intel just got done with their own Mobexit and despite continued interest in baseband modem chips just makes a bad fit.

    Stranger things have happened, but any of those four would upset the competitive balance of power in the ARM ecosystem should they move in. TSMC would make more sense since they service the majority of ARM’s customers and are also viewed as neutral. There is a possibility of another party – I’d suggest Amazon, who is trying to be in the chip business and covets the IoT, would be a great fit but would also have to venture into debt. So far, none of these names have moved past anything but wild speculation, and all might run afoul of UK government scrutiny.

    According to ARM, they are full speed ahead with SoftBank. The attractive part for ARM is a commitment to remain an independent operation, keep their Cambridge HQ and their existing management team, and add 1500 engineers in the UK. Everyone including Son is after the IoT, and it’s a big opportunity long term, but for SoftBank the 5G infrastructure play with ARM server chips inside may be even more important.

    In fact, it might be the formula to fix Sprint, reducing their infrastructure rollout costs and getting their next-generation network up and running quickly. It gets even bigger if Sprint figures out how to take over T-Mobile, a deal declared dead in 2014 but possibly being reanimated soon given a relentless need for more spectrum to remain competitive.

    Several analysts and editors are clueless on where could ARM possibly put 1500 engineers – assuming these are all chip designers is a narrow view, almost everybody missed the point. I can see ARM taking a much bigger role in software, extending mbed OS and OPNFV efforts for instance, and moving farther into deep learning and machine vision. Increased open source efforts from ARM for 5G could dampen any concerns from AT&T and Verizon about Sprint having an unfair advantage under a SoftBank umbrella. Add to that a play in IoT networks with something like NB-IoT, or the UK darling Weightless and a host of alternatives, and adding a lot more ARM engineers looks a lot more appealing.

    Then, there’s China. Robin Saxby worked very hard to open Asia for ARM, shuttling back and forth to Japan for a technology seminar binge before finally securing Sharp as the third ARM licensee in March 1993. Korea was not far behind, with Samsung signing a huge ARM license in May 1994 – aimed directly at beating the major Japanese electronics firms to the DVD player market. China’s nascent semiconductor market took another decade to adopt ARM, but now features a growing roster of ARM licensees.

    What happens in the Asia geopolitical competition, with ARM suddenly non-neutral in that context? Some analysts are saying China may turn elsewhere to avoid supporting Japanese technology. Speculation is rampant that the door may be now open for RISC-V architecture in China, and MIPS architecture is still strong there. However, given the relationship between SoftBank, Alibaba, and Tencent, it seems SoftBank has figured out how to do business with the Chinese. In a world moving to IoT and 5G, China needs all the help it can get.

    I initially agreed with Hermann Hauser, the man who inspired the teams who created ARM architecture, when he said this is a sad day for British technology. I have to believe ARM is out trying to smooth nerves right now, globally, and after giving it some thought I see very few feasible competitive bids out there. SoftBank has to hold up their end of the bargain, particularly the 1500 incremental engineers and not succumbing to a huge debt load or flipping ARM after a couple of years. I know Hauser and others have a great deal of pride in what they helped create, and he’s seen the carnage a disinterested parent company (Olivetti) created for ARM – we can only hope that’s not repeated.

    ARM’s mission, as with any public company, is to generate shareholder return. With mobile returns on the decline and IoT still not adding up to huge dollars, adding 1500 engineers at ARM would be out of the question near-term without a significant external capital infusion. SoftBank can help bolster IoT investments while protecting its stake in 5G infrastructure, and help ARM thwart Intel and RISC-V and others in the process.

    For the backstory of ARM and its history, particularly the early days with Hauser and Saxby as principals, I humbly suggest my book “Mobile Unleashed”.