GPU powerhouse Nvidia continues to try to keep its plans to acquire British chip supplier IP Arm Ltd. alive, despite continued concerns from several governments around the world.
In its latest action, Nvidia filed a 29-page response to the UK government outlining a list of potential benefits from the proposed $40 billion deal. Nvidia’s response follows a November 2021 action by the UK’s Competition and Markets Authority (CMA), which unveiled its intention to conduct an in-depth 24-week ‘phase two’ investigation into the proposed merger. . This investigation follows the CMA’s earlier phase one investigation, which determined that more information was needed to fully investigate the matter.
Nvidia’s CMA response filing was first reported by SearchAlpha. The Nvidia report was released by the CMA for public inspection after it was received from Nvidia.
Nvidia’s filing argues the CMA was wrong to order a phase two investigation, saying the facts of the case so far support Nvidia’s position that the acquisition would be good for the chip market and for Arm.
“Arm’s vision that the [CMA] describes the decision – an entity that ignores its profit motive and has no competition – is a mirage,” Nvidia claims. “Arm is a private, for-profit company at a crossroads. After acquiring Arm several years ago, SoftBank increased Arm’s workforce, hoping to drive long-term growth in several markets, including data centers and personal computers, long dominated by Intel and x86. . SoftBank’s investment phase is over, and one way or another, SoftBank intends to exit Arm.
Prior to Nvidia’s merger proposal, Arm had already considered an IPO but rejected that option, Nvidia argued. “As a standalone company, Arm faces significant growth challenges,” Nvidia’s response said.
“Nvidia did not approach SoftBank to purchase Arm,” the Nvidia filing continued. “Nvidia is a strong supporter of the x86 ecosystem and has developed accelerated computing platforms for x86 PCs and data centers throughout its history.” Instead, it was SoftBank that approached Nvidia about such an acquisition, the company said.
“The parties realized that Nvidia would be uniquely positioned to help Arm create a new IP and develop a world-class ecosystem that could provide an alternative to x86, providing customers with more choice and growing markets around the world. whole”, continues the file.
The proposed transaction would “significantly change Arm’s incentives and opportunities”, giving the combined companies “all the incentives and the ability to significantly increase Arm’s R&D investments at all levels, rather than face the tough choices of where to divest and face additional pressures from customers and competition,” Nvidia’s response said.
Nvidia also challenges another premise of the CMA’s Phase One report, which argues that the acquisition would prevent competition in many downstream markets.
“The logic and conclusions of the decision are flawed,” Nvidia said in its response. “First, the decision ignores competition from Intel in all relevant markets. Antitrust law preserves competition – it does not give customers and competitors a veto over acquisitions. The decision appears to lose sight of this fundamental principle. , saying instead that if enough high-profile Arm customers oppose the deal, the deal must be anti-competitive and should be blocked. But even if some customers and competitors aren’t happy with Arm’s plans, the deal won’t not likely to prevent competition.
The deal, if approved, could have no antitrust implications for Intel, which has been the main processor supplier for more than 30 years, Nvidia argued. “Downstream customers have several Intel options – Intel not only offers its own processor designs, but now also x86 IP licenses for third parties to create their own custom processor and SoC designs. hundreds of RISC-V supporters as forever unable to compete with Arm No industry observer can seriously argue that Intel, AMD and other Arm competitors are so incapable that they can’t even compete with arm.
With or without the merger, Arm cannot hurt competition, Nvidia argued.
“Adversaries of Deal romanticize Arm’s past and ignore or denigrate Arm’s most powerful competition,” Nvidia continued. “But if Arm had market power, it would have tremendous revenue growth and be extremely profitable. Rejecting the prospect of any recourse, the [CMA] decision would not promote competition. On the contrary, it would prevent Arm from introducing competition into areas that have long been dominated by x86. The alternative outcome demanded by opponents of the deal would be a stand-alone, profit-maximizing company, with no guarantees regarding licensing policy or investments. This would likely mean less investment in the UK, less resources for Arm, less innovation and less competition globally.
In a written statement delivered to EnterpriseAI, Nvidia said it submitted the response to the CMA to share more information about its intentions.
“Our phase one submission explains that the transaction will help accelerate Arm and drive competition and innovation, including in the UK,” the statement said. “We continue to explain the benefits of the transaction and are working to address regulators’ concerns in phase two.”
Spokespersons for the CMA and the UK Secretary of State for Digital, Culture, Media and Sport could not be reached for comment on Nvidia’s response today.
Nvidia’s 15-month, $40 billion proposal to acquire IP chip supplier Arm has faced a series of potential regulatory hurdles since October.
In December, the United States Federal Trade Commission (FTC) filed an administrative complaint on December 2 in an attempt to block the blockbuster transaction. The administrative complaint, which will be heard by an administrative law judge in August 2022, alleges that if the merger is approved, it will give the companies “the means and the incentive” to stifle innovative next-generation technologies, including those used to manage the data. driver assistance centers and systems in cars, according to the FTC filing.
The FTC’s administrative complaint followed related actions by regulators for the UK government and the European Commission (EC), including the UK government’s phase two investigation announced in November.
On October 27, the EC unveiled its own “thorough investigation” into the proposed merger proposal, saying it wanted to determine whether the merger could stifle fair competition in the market. A first phase investigation by the EC determined that it needed more time to assess the proposed merger and its effects before issuing a final decision in the case.
The three agencies – the FTC, the UK’s CMA and the EC – say they are reviewing the proposed acquisition more thoroughly over various concerns, including how it could stifle competition, reduce innovation in the semiconductor industry and cause potential harm to the domestic market. Security.
Nvidia’s merger proposal for Arm was announced in September 2020 and has been met with controversy as well as praise. Several big tech companies, including Google and Microsoft, have strongly opposed the deal, raising repeated concerns about its negative effects on competition and pricing.
But in June 2021, three other chip companies – Broadcom, Marvell and MediaTek – backed the acquisition and began to publicly state that they viewed the move as one that could ultimately benefit their own businesses.
Nvidia’s acquisition of Arm was set up when Japanese technology investment firm SoftBank, which bought Arm in July 2016 in a $32.25 billion all-cash deal, chose to sell the company after hemorrhaging cash since the first quarter of 2020. SoftBank was looking to sell assets to raise cash after the company’s earlier bets on the rise of connected devices failed to pay off . The company’s Vision Fund, its artificial intelligence investment fund, suffered an annual loss of $13 billion in its fiscal year ending March 2020.
Acquiring Arm would strengthen Nvidia’s position as a major player in wireless and other markets, as it makes steady inroads into enterprise data centers. The graphics leader has released a stream of ever more powerful GPUs targeting machine learning and other AI workloads that now dominate enterprise data centers.
Comments are closed.