Nvidia Just Raised Its Supply Commitments to $390 Billion, Mostly for Memory. Micron Is the Memory Stock I’d Buy.

When Nvidia (NASDAQ:NVDA) reported on Nov. 17, its fiscal third-quarter revenue had more than doubled from the year before, to a record $115.8 billion. But for memory investors, the more telling number was how much the artificial intelligence (AI) chip leader has agreed to pay its suppliers.

Nvidia‘s supply commitments rose to $390 billion, up from $279 billion a quarter earlier. And the company said the increase was “primarily related to the multi-year procurement of memory supply.”

Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a “Double Down” signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same “Total Conviction” signal is flashing for a company 1/100th the size of Nvidia. Continue »

That’s good news for Micron Technology (NASDAQ:MU), one of the three big makers of the memory chips Nvidia buys. But Micron stock is around $702 as I write, about 42% under the record $1,213.56 it closed at on June 25.

I think the market’s underrating what Nvidia’s report says about Micron’s business.

A Micron sign in front of a Micron office building lined with palm trees.

Image source: Micron.

Nvidia is buying memory years ahead

Showing how fast its buying has picked up, Nvidia’s supply commitments have more than tripled over two quarters. They were at $119 billion as of April 26, climbed to $279 billion by July 26, and hit $390 billion by Oct. 25, the end of its fiscal third quarter. In both of the last two quarters, the company linked most of the rise to memory. What’s more, the money’s coming due sooner. Of the $390 billion, $203 billion falls in Nvidia’s fiscal 2028, which ends in January 2028. Three months earlier, that year’s figure was $87 billion.

In short, Nvidia is locking in memory long before it needs the chips, and much of this spending will probably land within about a year.

How much of it goes to Micron?

Nvidia doesn’t say how it splits the spending between its memory suppliers. SK Hynix and Samsung Electronics sell Nvidia high-bandwidth memory (HBM), too. (HBM is the stacked memory that sits next to AI processors.)

Micron doesn’t name its largest buyers, either. However, no single customer made up 10% or more of its revenue for its fiscal 2026, which ended Sept. 3, down from one customer at 17% the year before.

So Micron’s direct sales to Nvidia, whatever they are, were under a tenth of its revenue last year.

What Micron does say is encouraging. It’s working with Nvidia on what it calls the industry’s first custom HBM4E chip, designed for Nvidia’s next generation of graphics processing units.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *