On July 18, SoftBank Group of Japan announced that it would buy ARM Holdings, the renowned UK-based semiconductor intellectual property (IP) company, for a whopping 43% premium over the recent stock price. While much consolidation has taken place in the semiconductor industry in the last couple of years, this particular marriage is a surprise. However, its long-term outlook could be good, so long as “the bank” stays just that.

Stay the Course

ARM’s stable management and strategy, evolving processor offerings in tune with market changes, and increasing performance to reach into more demanding applications have allowed ARM to reach nearly every electronics application, vendor, and original equipment manufacturer (OEM) over the years. The success has been outstanding. Even the mighty Intel has failed to stem ARM’s market presence, despite a full-frontal attack. Meantime, once-significant competing architectures like SPARC, MIPS, Power(PC), ARC, and untold proprietary processors have faded into the shadows.

SoftBank is a different sort of holding company, holding a disparate variety of businesses—mostly telecom (including Sprint), Yahoo!Japan, and maybe even a baseball team. Little appears close to semiconductors and electronic equipment, although SoftBank once bought memory maker Kingston Technologies—later nearly suffocating it. It is hard to find positive synergies between SoftBank’s holdings and ARM’s expanding markets. While mobile phones gave the shot in ARM’s revenues it needed 15 years ago, telecom infrastructure and cellular service providers (SoftBank) have little in common with semiconductor IP design, manufacturing, or licensing that ARM hadn’t already figured out on its own.

SoftBank said (and ARM reflected this) that it will leave ARM’s executive team and culture in place, maintain its headquarters in Cambridge, and double the workforce in Britain over the next five years. This implies a reasonably “hands-off” approach by SoftBank.

This analyst believes ARM has done a superb job of running its business over all these years, and all efforts should be made to stay the course. Past dealings with ARM have never indicated that its activities were constricted due to a lack of funds. However, being given a large infusion of cash and the ability to hire 300 new people a year would allow much faster progress to be made in the following areas:

  1.  Addressing emerging applications;
  2.  Enhancing software and standards support (key to making the hardware attractive); and
  3.  Ramping up performance to serve the highest performance applications.

A 15%-20% increase in engineering staff can really shorten project completion time. On the sales and marketing fronts, the increase would reach more potential customers. Oddly missing, however, are statements about worldwide personnel increases; ARM’s markets are largely outside of the UK.

Auto IoT

Two applications have been mentioned by ARM and SoftBank: the Internet of Things (IoT) and semi-autonomous driving. Those certainly are the hot buttons of the year. IoT encompasses almost anything electronic a person can think of, all joyfully communicating data between themselves and servers or the cloud. Every electronic company in the world has IoT on its must-do list, but the reality of usefulness and universal interoperability is still as much as 10 years away.

While ARM is very well-positioned to drive IoT coordination (a good use of that new headcount), and certainly to sell processor cores into IoT devices, there is stiff competition. Intel sees itself as a driver of the IoT, at least driving from the top—the infrastructure end—but Intel has yet to make much progress in the high-volume “Things” side after woefully failing to make a dent in the smartphone market. Most other semiconductor companies are also addressing the developing IoT industry—with ARM-based components. But so far there is little cohesion between the numerous IoT announcements, so a real “everybody’s talking to one another” IoT is surely a long way off.

Self-driving cars involve extremely complex computing systems, often requiring aid from municipal infrastructures that do not yet exist. While these are great engineering projects to work on, a vehicle that can handle the myriad driving environments and situations it will encounter will require much more research, development, and testing on digital electronics, legal, and human behavior fronts. That sounds like a long-term project. However, there is indeed a high probability that most of the processors in such vehicles and supporting electronics will contain ARM processors. Therefore, ARM should certainly devote resources to encouraging development of Advanced Driver Assistance Systems (ADAS) and autonomous vehicles.

Downsides

Taking a company like ARMH off the stock markets and putting it in a barrel with assorted other businesses adds new difficulties to high-level management and accounting of the company. Individual investors no longer can place their bets directly on their favorite company.

The executive ranks at ARM have many years with the company and often have technical backgrounds—vital in such a critical high-tech company. Loss of these leaders following a change of ownership, direction, and strategy can deal a serious blow to the organization’s success. Promises that executives will remain in place are hard to fulfill, especially when the individuals have stacks of cash flashed at them. Guidance from new management with little applicable experience can drive off accomplished execs in frustration. One-year and transition commitments only delay the ultimate departures without correcting the problem.

ARM has been a real technology feather in the United Kingdom’s hat the last decade or two. In fact, former CEO Robin Saxby was knighted. Selling this (crown) jewel to an organization in Japan may take some of the polish off, even if the business is still firmly rooted in proud Cambridge.

Many semiconductor companies that were taken private the last decade or so did not fare so well in the end, whether or not they were later reintroduced on the public markets. Most of those were struggling when they were absorbed or taken private. ARM is an interesting contrast, having managed well in the (stock) markets, in spite of being a curious IP company rather than the normal semiconductor company, and gaining market share—it is now on top of its game. How ARM looks in five or 10 years after this transition will be a unique case study.

A Positive Direction

An infusion of funds to strategically build the business is always welcome at any company, but any strings that are attached to those funds can really sour the fruit. Hopefully, ARM will be able to utilize its new finances and personnel to accelerate the rate it grows its markets: beefing up efforts in high performance computing, as well as the IoT and longer-term autonomous driving. Running faster than the competition and leading your customers with advanced technology is the route to success in the semiconductor business. A little extra cash can always help.

Note: Simon Segars, present CEO at ARM, gives a positive, well-spoken, realistic outlook for his company here.