Americas

  • United States

Nvidia-Arm merger faces regulatory, political, legal hurdles

News Analysis
Sep 15, 20205 mins
Computers and PeripheralsData Center

While Nvidia says it won’t interfere with Arm Holding’s chip-design business, its competitors and regulators may be skeptical

slow development red tape bureaucracy bureauocratic
Credit: track5 / Getty Images

Nvidia’s planned $40 billion takeover of chip-architecture firm Arm Holdings is not your typical merger. Oftentimes in a merger it’s one company taking over a weaker competitor that it has vanquished, something Nvidia knows all too well. Over its history, Nvidia has purchased several competitor GPU makers, most notably 3DFX in 2000.

But here, the situation is different. First, the two companies don’t compete. Nvidia was a licensee of Arm chip design with its Tegra processor aimed at smartphones and tablets—and a rare failure for Nvidia as it never really caught on.

Arm was purchased in 2016 by Japan’s SoftBank and the only reason it sold the company is because SoftBank is plagued with debt and needs some cash. It is also looking to sell its stake in T-Mobile, which it inherited through its ownership of Sprint, which merged with T-Mobile earlier this year.

The result is a merger of two equals who can hit the ground running when the merger is complete. “It really an accretive deal,” said Jon Peddie, president of Jon Peddie Research, who specializes in the graphics market.

“On the call with [Nvidia CEO] Jen-Hsun [Huang] and [Arm CEO] Simon [Segars} with analysts, the thing they kept pointing out is how complimentary everything is with the two companies, how Nvidia can expand the offerings to their customers as well as Arm customers. The implication is Nvidia might license their proprietary GPU stuff and make that available to Arm’s developer network,” Peddie said.

Jim Feldhan, president of Semico Research, echoes the notion that Nvidia just might sell its GPU technology through Arm channels. “They see Arm as an opportunity to sell their IP to other semiconductor or system houses,” he said. “They can sell to selected companies where they would believe they wouldn’t have the opportunity to sell it themselves.”

So that leaves Intel and AMD out. They’re both Arm customers that compete with Nvidia.

Feldhan thinks if Nvidia does license its tech, it won’t be the brand-new Ampere line but older IP. “I wouldn’t imagine they will sell their flagship product, but they have legacy product that would do well in other products. It allows them to reuse IP that they have already developed that is not on their front line but can be used in other apps,” he said.

Arm does have a GPU sold under the Mali brand name, but Peddie says it’s “a different animal. It serves the low-power, low-performance market for cheap phones. Nvidia was never going to get into that business.”

Any objections?

Arm has more than 500 licensees, which, in addition to AMD and Intel, includes giants like Apple, Qualcomm, and Broadcom. They or any other licensee can throw a monkey wrench into the deal with an anti-trust objection. Plus there are national interests that will be expressed by regulators in the U.S., the always-contentious European Commission and China, which scuttled a proposed merger between Qualcomm and NXP in 2018.

Indeed, no sooner had the Nvidia-Arm deal been announced than a group called Save Arm led by Arm co-founder Hermann Hauser popped up, imploring British Prime Minister Boris Johnson to stop the deal.

“There’s no incentive for Nvidia to upset the business plan Arm has executed on for 25 years. It’s a profitable plan that has worked, and Nvidia would be stupid to mess around with it—and I think we can agree Nvidia is not stupid.”—Jon Peddie, analyst

But Peddie thinks the management can calm customer concerns at the very least. “Once Jen-Hsun and Simon speak to their customers, those customers will realize this is not threatening but will offer them even greater capabilities than they had,” he said.

“There’s no incentive for Nvidia to upset the business plan Arm has executed on for 25 years. It’s a profitable plan that has worked, and Nvidia would be stupid to mess around with it—and I think we can agree Nvidia is not stupid.”

On a very late-night conference call with the press, Huang said just that. “I think [customers] just have to hear from us, and they have to hear from Simon, and of course Simon and the team are bought into it. And they’re ultimately the most passionate about the openness and the neutral way by which they serve all customers,” he said.

Big plans for Arm in Cambridge

Arm is headquartered in Cambridge, England, and on the call Huang said he had big plans to turn the location into Nvidia’s largest design center in Europe and hire thousands of people. “This will be the biggest thing to hit Cambridge in a long time,” said Peddie.

Huang also outlined Nvidia’s intention to build a world class AI lab in Cambridge “because we have a great site there already with Arm. From this facility, we will house a magnificent supercomputer that’s based on Arm CPUs and Nvidia GPUs and Mellanox chips and it’s going to be the most advanced AI supercomputer the world’s ever known.”

Andy Patrizio is a freelance journalist based in southern California who has covered the computer industry for 20 years and has built every x86 PC he’s ever owned, laptops not included.

The opinions expressed in this blog are those of the author and do not necessarily represent those of ITworld, Network World, its parent, subsidiary or affiliated companies.