So far in the AI revolution, developers have mostly relied on Nvidia GPUs to train their models. But the California chip maker may not be able to retain its monopoly for long.
Among those eying Nvidia’s crown, 2 of the world’s foremost tech investors have got their sights on the AI chip market: Softbank founder Masayoshi Son and OpenAI CEO Sam Altman.
During a presentation to SoftBank Group shareholders last week, Masayoshi said: “What I am most interested in and preparing for now is the AI revolution.”
Describing his belief that “super-intelligent AI machines” will surpass human intelligence by a factor of at least 10,000, he proposed a new AI chip venture to help position the firm at the center of the sector’s growth.
Codenamed Izanagi, Masayoshi’s plan is to leverage SoftBank’s 90% stake in Arm Holdings, a UK-based CPU manufacturer that has increasingly pivoted to the AI chip sector.
SoftBank itself intends to put up $30 billion to fund the new venture, Bloomberg reported.
“We went into defense mode and have over 4 trillion JPY on hand,” Masayoshi told the company’s annual shareholder meeting. But now, he said, “the time to shift back into offense mode is approaching.“
On top of the bank’s input, the plan is to raise an additional $70 billion from investors in the Middle East, where SoftBank has deep ties with the sovereign wealth funds of Saudi Arabia and the United Arab Emirates (UAE).
However, Masayoshi isn’t the only would-be silicon mogul courting institutional investors in the region.
According to recent reports, OpenAI CEO Sam Altman is also in talks with Emirati Investors to raise funds for a new AI chip manufacturer.
In fact, Altman’s ambition appears to be even greater than Masayoshi’s, and the company he has proposed may ultimately require between 5 and 7 billion dollars.
That figure would make the venture more valuable than any company in existence today. But Altman doesn’t just want to rival Nvidia. His plan envisions transforming the global landscape of semiconductor manufacturing from the ground up, reorganizing the established hierarchy for a new era of AI.