Eric Schmidt, who headed Google for ten years, during a conversation with Stanford students, said that not being an investment consultant, he sees an obvious play on the stock market, which did not involve buying shares of his former employer.
In a video posted by Stanford this week but then removed, Mr. Schmidt said that major technology companies are planning increasingly large-scale investments in Nvidia-based artificial intelligence data centers. According to him, the cost of construction of these functional facilities may amount to $300 billion.
Eric Schmidt also noted that he communicates with representatives of large companies who say that they need $ 20 billion, $ 50 billion, and $100 billion. According to him, most of these expenses are going to Nvidia. The mentioned company specializes in the development of basic artificial intelligence chips for data centers. For three consecutive quarters, Nvidia’s revenue has shown growth of more than 200%. This summer, the market capitalization of this company exceeded the $3 trillion mark.
Google has developed chips called Tensor Processing Units (TPUs). These products can potentially compete with Nvidia processors, but so far there has been no corresponding strengthening of the market position of the Internet search giant, which intends to benefit from the so-called artificial intelligence boom.
Eric Schmidt’s statements, made during a conversation with Stanford students, very unequivocally declared his vision that Nvidia is currently the largest beneficiary of the popularity of AI. In a comment to the media, he noted his request to take down the above-mentioned video of this meeting because he had misspoken about Google’s lax work culture in the session. Eric Schmidt headed the specified company from 2001 to 2011.
Nvidia is currently seeing an increase in demand for its products from cloud companies and leading developers of artificial intelligence models. Speculation is circulating on Wall Street that the company’s customers’ spending on machine intelligence infrastructure may be excessive. Nvidia is scheduled to release quarterly results, which will largely reflect the current state of the artificial intelligence market, on August 28.
