Arm and Nvidia. Heads or Tails

Acquisition approval rests on a knife-edge

Eight months have now passed since Nvidia announced its intention to acquire Arm from SoftBank for $40 billion. It’s little surprise that progress toward completion has been slow, complicated by regulatory review and antitrust concerns. Nvidia stated at the start that it would be an 18-month process, but the slow progress so far means that if the acquisition is to close at all it could drag on for much longer. The outcome remains a coin toss.

To say this a complex transaction is an understatement. There’s a myriad of factors at play that will ultimately determine if the transaction proceeds. We have consistently taken the view that the continued independence of Arm, and preferably as a public company again, is in the best interests of its ecosystem of licensees and partners. However, the sum agreed between SoftBank and Nvidia and a lack of comparably attractive options mean there’s a stark commercial argument despite opposition to the deal (see Opposition Mounts to Nvidia’s Bid for Arm).

In recent weeks there’s been a series of developments that raise hurdles on the path to completion. The first was a decision by the UK government to intervene and examine the transaction on grounds of national security. This will see the UK’s Competition and Markets Authority prepare a report on the implications of the transaction by 30 July, and is in addition to the review announced in January to determine whether the deal could lead to Arm withdrawing, raising prices or reducing the quality of its intellectual property licensing to Nvidia’s rivals.

Whether a national security dimension is likely to derail the deal is difficult to predict. The strategic importance of semiconductor design and manufacturing means such deals are given a high degree of scrutiny, as President Trump’s intervention in Broadcom’s attempt to acquire Qualcomm in 2018 illustrates. On the face of it, the US–UK dynamic at play in the Nvidia–Arm deal would suggest this is less likely.

However, a bigger consideration is going to be China’s role in the regulatory clearance process. China has voiced reservations about Arm moving under the control of the US government — a significant worry given the state of relations between the US and China and the strategic importance of Arm technology and the role it has played in establishing China’s dominance of consumer electronics.

The clearance process is a series of negotiations, so national security concerns could hinge less on the commercial details and more on the concessions China might seek before it gives the green light to the deal. The challenge here is that China is often the last regulatory body to grant clearance, so anticipating China’s demands is difficult. This complexity could result in further delay to the process as regulators look to one another to commit first.

The Chinese element of the transaction is also complicated by the refusal of Arm China CEO, Allen Wu, to resign after being ousted by the board.

The political climate in Washington and the growing scrutiny that large-scale acquisitions now face are further obstacles. They are by no means specific to Nvidia — a bill proposed by Senator Amy Klobuchar promises widespread reform to antitrust rules that would make acquisitions of any size and profile significantly harder. The impact would be far-reaching, with potentially negative implications for the start-up economy. But that’s a blog for another time. Social media platforms are the focus of such reform, but all merger and acquisition activities, particularly those of such value and significance, are now more squarely in the spotlight.

Nvidia’s recent launch of Grace, its first Arm-based CPU, also adds ammunition to the argument that it doesn’t need to acquire Arm at a cost of $40 billion in order to compete effectively in data centre CPUs. More pertinently, Nvidia has repeatedly said it doesn’t compete with Arm licensees. Grace undermines this claim, because a string of Arm licensees are eyeing the same opportunities in data centre silicon. Rivals are concerned that Nvidia’s control of Arm would enable it to more closely tailor the architecture for its own advantage.

The acquisition will undoubtedly draw further scrutiny by the US Federal Trade Commission and other regulators. It also highlights that there is more complexity to consider than in a standard vertical acquisition of two companies providing different supply chain functions. Whether the deal ultimately closes remains uncertain but what’s clear is that delay and uncertainty are anything but positive for the Arm ecosystem.