Okay, this is getting weird. Not just the leasing which is less profitable for Nvidia but the term of the lease which is being mooted at four or five years, longer than the economic life of the chips.
A quick hit from TipRanks, September 23:
Microsoft-backed (MSFT) OpenAI and Nvidia (NVDA) are in talks over an unusual way to structure their new AI data center partnership, under which the ChatGPT maker would lease Nvidia’s AI chips rather than purchase them, The Information’s Anissa Gardizy and Sri Muppidi report, citing two people who spoke to executives at the companies about it.
That's it, that's all we have since dropping The Information subscription.
Nvidia seems to be getting creative in how it deals with customers, here's Tom's Hardware on a different deal, September 5, referring once again to The Information:
Details are fuzzy.
Nvidia, the company powering essentially the entire AI boom, is now looking to rent back its own chips. According to The Information, Nvidia has now signed a deal with Lambda, a cloud service provider in which it has already invested, to lease back some of its own GPUs over a period of time.
The details of the purported agreement are unclear, but reports suggest a four-year lease for 10,000 GPUs, valued at approximately $1.3 billion, followed by a separate $200 million deal for an additional 8,000 (likely lower-end/older) chips. These are Nvidia GPU servers that Lambda once bought from the company, and now the chipmaker wants to rent these back. This contract will make Nvidia, by far, the biggest client under Lambda's belt....
....MUCH MORE
Both these stories tie back to something we were thinking earlier today:
"Nvidia Stock Pulls Back From Record High. Wall Street Has Questions About OpenAI Deal." (NVDA)
Besides the whole ouroboros thing with Nvidia investing money that will come back to Nvidia as chip sales. Something we pointed out in November 2024's "Nvidia Reportedly Weighing Investment in Musk’s xAI" (NVDA; ELON):
Although the A.I. biz is not yet incestuous it does have a bit of an ouroboros vibe to it....
—from 2018's Unicorns Backing Their Own VCs? Welcome to Peak Tech
Something that Bernstein is also addressing. From Bloomberg via The Edge, Singapore, September 24:
Nvidia’s massive OpenAI deal fuels ‘circular’ financing concerns
Three years after OpenAI and Nvidia Corp helped kick off the global artificial intelligence frenzy, the two firms are joining forces to pave the way for a more costly phase of development with a deal that’s quickly revived fears of an AI bubble.
Nvidia on Monday said it will invest as much as US$100 billion in OpenAI to help the ChatGPT maker support a massive build-out of data centres equipped with Nvidia’s chips – a deal that some analysts say raises questions about whether Nvidia is investing heavily to prop up the market and keep companies spending on its products.
“The action will clearly fuel ‘circular’ concerns,” Stacy Rasgon, an analyst with Bernstein Research, wrote in an investor note after the deal was announced.
Those concerns have followed Nvidia, to varying degrees, for much of the AI boom. The chipmaker participated in more than 50 different venture investment deals for AI companies in 2024, and is on pace to top that number this year, according to data from PitchBook. Some of these companies, which include AI model makers and cloud providers, then use that capital to buy Nvidia’s expensive graphics processing units.
But the size of the OpenAI investment “appears to dwarf all the others,” Rasgon said. The deal will “likely fuel these worries much hotter than what we have seen previously, and (perhaps justifiably) raise concerns over the rationale behind the action,” Rasgon wrote. Nvidia has said the investment in OpenAI will not be used for any “direct purchases” of Nvidia products, Rasgon added.
Other large tech companies, including Microsoft Corp and Amazon.com Inc, have also made strategic investments in top AI startups intended to drive business to their cloud computing offerings. But Nvidia occupies a unique spot in the AI ecosystem by dominating the market for advanced chips that are essential for training cutting-edge AI models....
....MUCH MORE
