The Long Arm of Competition law? NVIDIA/Arm : can the deal overcome innovation concerns?

The European Commission announced that it has opened an in-depth Phase 2 investigation into the proposed acquisition of Nvidia of chip designer Arm, which is currently owned by Softbank. The Commission has cited concerns that the deal could stifle innovation and the review will have important implications for other semiconductor and tech sector deals more generally, including those subject to parallel review by major global regulators.

While Nvidia develops and supplies semiconductors for various applications, including in datacentres, IoT automotive and gaming, UK-based Arm licenses out IP for processing units, in particular to semiconductor chipmakers and Systems-on-Chip  developers. The parties announced the deal back in September 2020, stating that by “uniting NVIDIA’s AI computing capabilities with the vast ecosystem of Arm’s CPU, we can advance computing from the cloud, smartphones, PCs, self-driving cars and robotics, to edge IoT, and expand AI computing to every corner of the globe.”

Nvidia and Arm do not compete with each other – they operate at different levels of the global semiconductor supply chain – meaning this is a so-called ‘vertical’ deal in competition terms. Vertical deals were historically seen as more benign from a competition perspective in the absence of one party having market power because they do not reduce direct competition between the parties and can lead to efficiencies like the removal of double mark ups. However, the Commission is concerned that Arm has significant market power on the market for the licensing of CPU IP for use in processor products and, therefore, that the combined Nvidia/Arm would have the ability to restrict or degrade access to Arm's technology by Nvidia’s competitors. It will also examine whether the deal could harm innovation by (i) making licensees of Arm IP reluctant to continue sharing commercially sensitive information with the combined entity because they are competitors of Nvidia or (ii) refocusing Arm's R&D spending on products that are most profitable for Nvidia downstream, to the detriment of players heavily relying on certain Arm IP in other areas.

The Commission has until 15 March 2022 to issue its Phase 2 decision. And the deal is also subject to review by other major regulators: it has been subject to an in-depth Second Request review in the US since last December, the UK CMA recommended in August that the deal be referred to Phase 2 (following intervention by the UK government), and in China, the clock has not yet even started running because the filing has not been formally accepted by the State Administration for Market Regulation (SAMR).

If any of the regulators ultimately identify concerns, there is no obvious structural cure to the problem, as it does not result from a horizontal overlap between the parties. While the European Commission has been willing to accept complex behavioural remedies in tech sector deals, including Google/Fitbit, the CMA’s Chief Executive Andrea Coscelli said in February that the behavioural remedies accepted by the Commission to clear Google’s acquisition of Fitbit would likely have been rejected in the UK (as they were in Australia). The CMA later released a joint statement with the German and Australian competition regulators arguing that the “increasing complexity of dynamic markets and the need to undertake forward-looking assessments require competition agencies to favour structural over behavioural remedies.” Therefore, even if the Commission concludes it can remedy any competition concerns through the use of behavioural remedies at Phase 2, other regulators (notably the CMA) are likely not to be willing to accept such remedies should they find concerns.

Against the backdrop of the global supply chain issues and continued scrutiny of semiconductor deals, it will be interesting to see where the major global regulators come out on this deal. Will there be coherence or will the regulators diverge on the substantive analysis, including concerns around innovation, and any remedies to address such concerns? If one of the regulators decides that the deal raises concerns that cannot be solved, it’s game over for the whole deal.