• The NVIDIA-Arm deal has interesting technological potential, but will likely chill competition
• Regulators worldwide are viewing big tech deals with an increasingly skeptical eye
In the ongoing saga of NVIDIA’s proposed purchase of the UK-based silicon design firm Arm Semiconductor Ltd. regulators have stepped in to stop the deal. Arm develops the architecture of the ARM processor, and then licenses it to other companies for use in their designs. ARM-derived processors have become extremely popular, appearing in almost every modern smartphone design, thousands of other proprietary, servers, and probably most famously as the latest CPU architecture for the Macintosh line of computers from Apple. Amazon’s AWS service has servers that AWS developed that use ARM architecture. In short, ARM is essentially everywhere and only Intel’s x86 architecture has had more success. ARM is the first processor architecture to get anywhere close and is considered vital in the technology marketplace.
NVIDIA for its part, is best known as the maker of graphic processing units (GPUs) for the consumer market, but has been investing in the data center market for years, including the purchase of networking firm Mellanox. GPUs have become important in the data center marketplace particularly in areas such as AI, digital twins, and AR/VR. NVIDIA is particularly invested in AI and sells a large assortment of data center products. NVIDIA, through its Mellanox subsidiary, also takes part in the data processing unit card (DPU)/network accelerator card market.
Last year, NVIDIA made an offer of shares plus cash to the tune of $40 billion. With the price of NVIDIA’s stock going up, the deal is now likely worth north of $50 billion. It has been previously announced that the UK Competition and Markets Authority has opened a formal inquiry and Chinese regulators are conducting their own review of the proposed deal. Now the U.S. Federal Trade Commission has filed suit to block the deal, citing concerns around NVIDIA controlling Arm, when many of NVIDIA’s data center competitors have invested in designed based on the ARM architecture. The FTC believes that NVIDIA’s acquisition of Arm Semiconductor will harm competitors and reduce competition in the marketplace. There is also concern that NVIDIA will gain access to competitor’s secrets contained in the licensing agreements that Arm has with those companies.
NVIDIA has claimed from the beginning that it will maintain Arm Semiconductor’s neutrality and open licensing model and has reiterated its intent to expand Arm and seek ways to expand Arm’s technology and reach. However, merger promises have a way of not working out the way regulators believed. Merger promises are seldom enforced afterward, with these immense firms having the resources to tie regulators up in knots until enforcement becomes moot. If you add in the current belief that tech companies have become too large and have too much influence, it doesn’t look good for the NVIDIA and Arm deal.
Arm has created an enormous and highly competitive market of chips that use its architecture, specifically because it does *not* compete with its customers. Arm does not sell ARM-based chips, it is a technology development and licensing firm. This puts it in a neutral position with everyone, including NVIDIA, to help them succeed using Arm’s designs. While its tantalizing to think about the technology that could come out of an Arm and NVIDIA deal, the chilling effect on competition likely isn’t worth it, especially considering that ARM chips are the only non-x86 chips out there that represent a threat to the x86 hegemony.
Considering the rising tide of regulator concern from multiple countries, this deal is likely to fail. The proposed deal has quietly alarmed NVIDIA competitors who are no doubt exerting what pressure they can to block the deal. The political atmosphere has also turned sour against big technology companies and this deal would be one of the biggest in the semiconductor industry. Its hard to argue that Arm will be *more* successful than it already has been and the downsides of the deal would not seem to be in the long term interest of the public or the tech sector, despite early enthusiasm for the deal.
Enterprises and IT professionals shouldn’t worry too much about the deal right now. Any plans for the next few years are not likely to need changes, even in the event of the deal going through. It will be interesting and enlightening to see how regulators rule on the deal and by what reasoning they confirm or deny. The result of the NVIDIA – Arm deal may be a blueprint for how regulators will rule on future technology firm mergers.