If anything, it now seems to be one of the most underestimated acquisitions in the history of tech: when nVidia announced that it’s buying ARM from Softbank for 40 billion dollars, all the way back in September 2020, it was considered to be a big deal but not a market-disrupting one. Everyone was focused on what nVidia would do to enter new markets with ARM’s know-how in low-consumption processors and what ARM could do with nVidia’s know-how in graphics processing. Consumers, as well as enterprise customers, were supposed to benefit from synergies between the two companies. But then a few important things came into focus… and now the deal is under scrutiny in both Europe and the US, with China and the UK following suit.
And that — despite the initial enthusiasm about this merger — is actually a good thing.
So what happened? A number of things, but mainly this: ever since the acquisition’s announcement, we’ve all been commenting on it under the impression that nVidia would allow ARM to operate in the “neutral” way it always has, but European and American regulators are now highly skeptical of that scenario. ARM traditionally designs the microprocessors it licenses as a “blueprint”, i.e. in such a way that other manufacturers can base their own chips on the same design but customize it as needed. That’s why Qualcomm’s, Apple’s, Samsung’s, and other manufacturers’ processors in smartphones and tablets all offer different performance and quite different features, despite being based on ARM’s architecture and general microchip design.
What the European Commission (who has opened an investigation on the acquisition) and the FTC (which outright sued nVidia in order to block the deal) are now afraid of is that ARM will not be able to retain its independence and neutrality towards other manufacturers under its new management. Sooner or later, regulators think, nVidia will try to either gain access to ARM’s designs before everyone else or not allow other manufacturers to have access to those microchip blueprints at all. Needless to say, synergies between the two companies could also end up being exclusive to nVidia’s products (since ARM does not actually bring to market any commercial products of its own). This, all regulators agree, will offer nVidia a number of unfair advantages and will ultimately stifle competition, slowing down the progress of microprocessor design as a whole.
There is one other thing that brought into sharp focus how harmful a monopolistic use of ARM’s designs by nVidia would be for the tech market globally: Apple’s success with the M1 line of processors (which is based on ARM’s architecture) during the last 12 months. These chips are so good for tablet, laptop and desktop products — balancing performance, advanced functions and power consumption in a shockingly impressive way — that most media outlets are now officially calling these smarter, modular, more agile, more eco-friendly processor designs “the future” compared to e.g. Intel’s monolithic and power-hungry chips. It’s no wonder that Intel has started slowly shifting to ARM-like designs too (big.LITTLE architecture), while AMD and others are expected to do the same at some point.
In other words: ARM’s microchip architecture is now more important than it was back in fall 2020 — so the company’s acquisition is a much more complicated matter too. Even nVidia itself admits that the very same mix of cash, stock, and equity that the deal cost it back then has almost doubled in value, exceeding the 75 billion dollar mark. With European, British, American, and Chinese regulators all opposing the deal, it’s highly likely that it will not be completed by March 2022 as nVidia hoped. This scenario would push the cost of the acquisition even higher.
What will be interesting to see, then, is how nVidia itself plans to proceed. Will it try to give assurances regarding the operating independence of ARM? How would it do that in a manner that’s convincing? Will it accept to include terms that deprive the company of any competitive advantages compared to other manufacturers? What would it be buying ARM for, then, if not for privileged access to that company’s output? It’s obvious that nVidia will not be able to complete the acquisition without regulatory approval, but it’s quite difficult to imagine how it could get that without the merger itself becoming less enticing (and much more expensive to boot).
Not a done deal by any means, then, but definitely one worth following closely. What many of our smartphones, tablets, and computers of all kinds will be able to do in the future — and how much they will cost — pretty much depends on this deal’s outcome, after all.