Beyond training, Chinese tech companies also use south-east Asian data centres to service their overseas customers, as Alibaba and ByteDance seek to grow their share of the global cloud computing market. Chinese companies are also expanding access to data centres in other regions such as the Middle East.

Nvidia , the world's most valuable company, has gone on the defensive against skeptics of its $4.5 trillion valuation, down from a historic $5 trillion, by waging an information campaign on Wall Street and social media.

This year’s revenues are anticipated to be around $215 billion, with the figure expected to surpass $300 billion next year.

The fifth metric is earnings per share. On a non-GAAP basis, third-quarter earnings per share came in at $1.30, up 60% year-over-year and 24% versus the second quarter. Generally accepted accounting principles (GAAP) earnings per share rose 67% year over year to $1.30 as well.

Top Chinese firms are training their artificial intelligence models abroad to access Nvidia's chips and avoid U.S. measures aimed at curbing their progress in advanced technology, Financial Times reported on Thursday.

Nvidia’s customers have a big incentive to explore cheaper alternatives. Bernstein, an investment-research firm, estimates that Nvidia’s GPUs account for over two-thirds of the cost of a typical AI server rack.

By investing in Nvidia, SoftBank gained exposure to the infrastructure portion of the AI pipeline, including Nvidia's GPUs, networking equipment, and AI chips in data centers. With its huge commitment to OpenAI, SoftBank is gaining exposure to the software portion of the AI pipeline (applications, models, enterprise tools, etc.).

For as long as Nvidia Corp. has dominated the market for artificial intelligence chips, customers have made clear they’d like to see more competition. It turns out one of the most formidable alternatives may have been hiding in plain sight.