The deal also could unsettle some Arm customers, who compete with Nvidia on specialized AI and graphics chips. That might help Intel or provide a boost for a nascent open source chip architecture called RISC-V.Nvidia said it would continue licensing Arm’s designs to customers, “while maintaining the global customer neutrality that has been foundational to its success.” It said it would also provide access to Nvidia’s technology to Arm’s customers.Nvidia said Sunday it will buy Arm for $21.5 billion in stock and $12 billion in cash. SoftBank, the Japanese conglomerate that currently owns Arm, may receive a further $5 billion if Arm meets certain financial targets, and Nvidia said it would issue $1.5 billion in stock to Arm employees.
Nvidia, based in San Jose, is best known for selling specialized graphics chips for PCs and game consoles. But its chips are also used within data centers to train AI algorithms. The company has benefited from a spike in demand for gaming hardware and cloud computing due to the pandemic; it is the world’s most valuable chipmaker, with a market capitalization of about $350 billion.Arm, located in Cambridge, England, licenses designs known as instruction-set architectures for central processing unit (CPU) chips, an alternative to those made by Intel , to companies including Apple, Google, Amazon, and Huawei. The company’s designs are found in about 90 percent of the world’s smartphones. It was acquired by SoftBank for $32 billion in 2016.
But while tech giants like Google and Amazon and Facebook have pushed major advances in the development of AI in purely digital contexts—getting computers to recognize objects in images, for example, by having humans label those objects first—robots have remained fairly dumb as researchers have focused on getting the things to move without falling on their faces.
The Secret to Machine Learning? Human TeachersThe two companies are dominant in AI and mobile, but Intel remains king of the data center. Only recently have some companies, including Amazon , begun adapting Arm’s designs to make more efficient new data center chips.Data center sales are a fast-growing part of Nvidia’s business, generating $3 billion in revenue last year, up from growing from $830 million in 2016. In April, the company acquired Mellanox, an Israeli company that makes hardware transmitting data between chips in cloud computing systems, for $6.9 billion.In an interview with industry analyst Patrick Moorhead, Nvidia CEO Jensen Huang said adding Arm to Nvidia would help sell more Arm chips to data centers. “What will change is the rate of our road map,” Huang said. “We know for sure that data centers and clouds are clamoring for the Arm microprocessor, the Arm CPU.”
It was a celebration of how exponential upgrades from the chip industry have propelled progress in technology and society over the past 50 years—and an argument that the party’s not over.“It's going to keep going,” said Jim Keller, a semiconductor rock star who joined Intel last year as senior vice president of silicon engineering, and a cohost of the event.
In coming years, more computing is likely to drift into the cloud, especially as more companies make use of AI for business applications and as 5G wireless networks create new possibilities for sharing data and applications.