The ARM-Softbank Delusion

So it may be helpful to point out what the problems are, so that in their next joint letter, the CEOs can say something of substance.

ARM got its start because Wintel dominated the PC industry and Jorma Ollila, CEO of Nokia which then ruled the mobile world, made it his aim to see that Wintel would not dominate the mobile market.

The Wintel experience of having one company manufacture all the processors in an important market segment was so painful that, when ARM came along with a processor which multiple manufacturers could make in competition with each other, the world embraced it.

ARM’s IP licensing model gave interoperability to the products in an industry segment without bestowing monopoly control on any one processor manufacturer.

Licensees trusted ARM to deal with them all equally in order for them to compete with each other on an equal basis.

And that worked well for 25 years until ARM was bought last month by telecoms operator SoftBank.

The key issue is: Will ARM continue to deal with its customers in an equal basis? And, just as important: Will ARM’s licensees continue to believe they will get fair and non-discriminatory treatment from ARM?

ARM’s massive stock of customer goodwill, built up over years of trustworthy trading, is on the line. If it starts to erode, the ARM business model won’t continue to work.

Those customers who have a choice – i.e. non-mobile customers like those making ICs for servers, IoT, MCUs consumer – may make other choices and, if so, will that stymie ARM’s growth?

Will ARM start to compete with its customers? For years analysts have speculated that ARM may adopt the fabless model, or a part-fabless model, and start fabbing it’s own cores. That would also destroy the business.

Then, one wonders, now that ARM is owned by a large telecoms company, will ARM’s designers be less motivated by the constant prod of competitive survival to keep on producing superlative cores. While chip companies feel the footsteps of competitors behind them all the time, telecoms companies behave more like utilities – secure behind their entrenched physical assets. Can, one wonders, a telecoms company make good decisions for a semiconductor IP company?

Furthermore, will UK universities and research houses be happy to work with ARM knowing that the results of the research will be owned by a foreign company? Particularly this applies to the mighty intellectual input from Cambridge University.

Cambridge’s intellectual contribution was a sine qua non while ARM was one of its progeny – it’s own flesh and blood. But now, absorbed by a giant foreign company, the automatic, unquestioning support of Cambridge may not be so forthcoming.

ARM’s aspiration, enshrined in its logo, was to be ‘The Architecture for the Digital World’. It was an expansive ambition in that it aspired to move from its base in mobile telephony into computers, servers, consumer and MCU.

Apart from MCU, it is has not been successful in expanding, but ARM has been patiently managed for the long haul and it has gradually strengthened its position in these markets. One wonders how patient a heavily indebted foreign entrepreneur will be. Softbank has $120 billion of debt.

Hermann Hauser who, back in 1990, organised the founding of ARM as a jv between Apple, VLSI Technology and Acorn backed by £1.75 million of capital, said: ‘I regret the fact we have lost our last independent hi-tech company with a global reach. There was no reason for change. It had £1billion of cash.”

26 years later that £1.75 million of initial capital has multiplied 13,714x to be worth £24 billion. It is one of the greatest financial success stories in the history of tech but, more importantly, ARM develped a key resource on which the industry could rely.

Bringing happiness to everyone is a nice thing, but destroying an asset which the world has found useful is a bad thing.

So c’mon you two CEOs, give us some facts about how this takeover will improve the ARM proposition.