(Reuters) – Nvidia Corp on Monday said it plans to build a server processor chip based on technology from Britain’s Arm Ltd, placing it in the most direct competition yet with rival Intel Corp and adding a layer of complexity to the antitrust review. Nvidia’s $ 40 billion deal to buy Arm.
Intel is the world’s largest maker of central processors for data center servers, but competition from Arm-based chips is increasing. With “Grace” server processors, Nvidia will be the largest chip company so far to challenge Intel in its core market.
Nvidia shares were up 2.6% and Intel shares were down 4% in midday trading after the news. Intel shares have risen more than 30% this year as new chief Pat Gelsinger outlines his strategy for dealing with years of manufacturing woes.
Arm’s technology powers the chip in most smartphones, but over the past few years it has made its way to data centers, with companies like Amazon.com Inc and Ampere Computing designing chips for servers with it.
Nvidia’s entry to the market could accelerate Arm’s breakthrough into data centers. Nvidia chips have traditionally been used as “accelerators” along with existing central processors from Intel, Advanced Micro Devices Inc or others, taking some of the computing work out of them and speeding up the overall computing system.
By building its own central processor, Nvidia is taking on Intel and AMD head-on for the first time. In a keynote address announcing the chip, Nvidia Chief Executive Officer Jensen Huang called the new server chip “the final piece of the puzzle” that will join Nvidia’s graphics and network chips to form “the basic building blocks of the modern data center.”
Santa Clara, California-based Nvidia says its “Grace” server processor, due for release in 2023, is designed to handle tasks such as training artificial intelligence algorithms. Once recognized for its graphics chips for enhancing video game play, Nvidia has focused heavily on artificial intelligence computing in recent years, helping its market cap overtake Intel to become the largest semiconductor company in the US.
Nvidia also said that the Swiss National Supercomputing Center will build a new supercomputer dubbed the “Alps” using the new “Grace” processor built by Hewlett Packard Enterprise Co. The computing system, which according to Huang will be 10 times faster than the best in the world today. supercomputers, will be used for tasks such as weather forecasting and physics research.
Nvidia officials say that the company’s accelerator chips will continue to support central processors from Intel and AMD even as they dive into the market themselves. Huang said that Nvidia will shift its focus, concentrating on making accelerator chips that work with Arm processors one year, and then chips that work with AMD and Intel chips next year.
Nvidia did not disclose which Arm technology would be used for the chip but said it would use Arm’s core “Neoverse” server chip technology due to be released in the future. Huang said it was Arm’s “off-the-shelf” technology.
Nvidia said in September that they would buy Arm for $ 40 billion.
Rival Nvidia has expressed concern that the company will use the potential of Arm’s holdings to gain early access to Arm’s technology as it becomes more prevalent in the data center market.
Arm said that Nvidia had the same access to its intellectual property as Arm’s other customer base and did not get early access to the technology.
Reporting by Stephen Nellis in San Francisco; Edited by Lisa Shumaker