Potentially more significant than the deeply unethical NSO Group hack or Apple’s new product plans for 2022, Nvidia’s attempt to acquire Arm seems unlikely to succeed, leaving the world’s most important silicon development company in limbo.
Who will buy Arm?
A very short list of contenders
First, a little background.
Nvidia surprised everyone when it agreed to pay about $50 billion to purchase Arm from Softbank. The problem with the move is that once competition regulators scrutinized the deal, they felt it was anti-competitive. The acquisition now faces opposition by the US Federal Trade Commission (FTC), which is suing to prevent the deal, and regulatory action in the UK and Europe.
The agency believes the deal will allow Nvidia to stifle competing next-generation technologies. “The FTC is suing to block the largest semiconductor chip merger in history to prevent a chip conglomerate from stifling the innovation pipeline for next-generation technologies,” FTC Bureau of Competition Director Holly Vedova said in a statement “Tomorrow’s technologies depend on preserving today’s competitive, cutting-edge chip markets. This proposed deal would distort Arm’s incentives in chip markets and allow the combined firm to unfairly undermine Nvidia’s rivals.”
David Bicknell, Principal Analyst at GlobalData, says:
“As the regulatory noise intensifies, the picture gets clearer: this bid will not get the green light from regulators.”
What can Softbank do?
Softbank doesn’t seem to have many options. Not only is the regulatory environment growing more challenging, but the list of potential acquirers is looking lean. It may choose to spin Arm off in a US IPO, or seek UK government support to keep it in the country. It is unlikely private equity investors will be interested, given the challenges Softbank faces cashing out.
Nvidia might be able to put together an alternative business plan designed to satisfy regulators and reassure competitors (including Apple), but to do so it will have to commit to some major compromises.
[Also read: What to expect from Apple in 2022]
That’s the business situation.
But the reality is that as a growing number of tech firms follow Apple to adopt Arm processor designs (including Microsoft and Qualcomm), uncertainty concerning the consequences of future ownership of Arm is at best non-trivial. At worst, given that it is impossible to see what Arm’s future will be, it is extraordinarily dangerous.
One of the primary arguments the FTC makes concerning Arm licensing is that Nvidia would gain access to sensitive information concerning licensees.
Like so many companies, Apple is a licensee of Arm, which means it very likely shares competitively sensitive information with Arm, which is seen as a neutral partner.
In the event Nvidia took possession of Arm, that neutral status would be in doubt, and the chip development company might become less likely to innovate in ways challenging to Nvidia’s interests. Nvidia could also gain access to historically confidential data from Apple and other competitors.
What about Apple Silicon?
I imagine danger is what Apple sees in all of this.
Mike Orme, Consultant Analyst at GlobalData, says:
“The elephant in the room is Apple. Despite relying on Arm instruction set for its foundational A and M series families of chips, Apple has kept quiet. It’s an open secret that Apple was offered ARM in 2020. It’s hard to think that Apple is deliriously happy about the prospect of the deal going through.”
Apple has raced to the top with the chips inside its iPhones and new Macs on the back of its work with Arm reference designs. And while it seems reasonable to assume the company is already exploring Arm alternatives, there aren’t many sufficiently robust to represent the future, yet. Though RISC-V could conceivably be one of them.
Acquisition is a possibility, though unlikely. In the event Apple chose to purchase Arm (or create a group to do so), it holds two cards:
1: It has more cash on hand than most other tech firms.
2: Arm could be considered cheap, given it is based in Brexit-battered Britain.
Apple would still need to reassure competitors – including Qualcomm – that it would not act in an anticompetitive fashion. Given Qualcomm has previously tried and failed to acquire Arm and given the current détente between the two firms perhaps they could find some way to work together?
In truth, the emerging competitive difference between the two will make that difficult, which sadly also implies that any attempt to purchase Arm by either firm will face opposition from the other. There’s also the challenge that may emerge from regulators in other nations in the event a US firm attempts to take control of the strategically important chip design company. That challenge could prevent other potential investors (Intel?) from moving into the frame.
So, what does this mean for users?
In the short term this means very little. We’ll continue to purchase Macs, iPhones, iPads, PCs, Surface tablets, Qualcomm-powered Arm-based smartphones and emerging smart home devices.
But in the medium term, this uncertainty at what is arguably the world’s most strategically important processor design company may limit future innovation.
If that happens, the many companies – including Apple – so dependent on that company today will find themselves in some form of limbo, as their competitive advantage depends on continued innovation around the core reference design.
If there’s a beneficiary here, it may be Intel, which must surely now hope to use this period of confusion to catch up with its own designs.
Apple’s teams, however, will be concerned now that they have proved the significant competitive advantage they can unleash by developing their own variants based on Arm reference designs.
The company will not want to lose the processor industry leadership it has been working towards for over a decade on strength of some corporate ownership struggle.
In the long-term?
I think Apple will need to secure control of sufficient silicon development technologies to build its own processors without Arm or anyone else. Even if it passes on Arm, this implies additional M&A activity perhaps as significant as the PA Semi purchase or acquisition of Intel’s modem teams.
Apple’s silicon teams will have been considering the possibilities since Softbank first began discussing its Arm divestment. If nothing else, the Apple Silicon team has proved its capacity for forward planning, which makes it likely they are already building a contingency plan.
Perhaps that plan is further along than we thought.
After all, Apple first introduced its own Arm-based chip design in 2010, which means Apple Silicon is now more than a decade old. Perhaps the best time to migrate to a new generation is when the current model is at the top of the league?