I have spent a long life watching people confuse activity with achievement. In the artificial-intelligence business, there is a great deal of activity. Nine-figure signing bonuses. Data centers the size of counties. Chatbots that talk back before you have finished your sentence. Most of it, I would gently suggest, is priced for perfection and built for disappointment.
So it is a pleasure, for once, to point at something in this industry that a plain old value investor can actually admire.
This week, buried in a routine SEC 13F filing, Nvidia disclosed what it has been quietly sitting on. As of June 30, the chipmaker held eight stock positions worth about $63.44 billion. Two of them make up better than 80 percent of the pile: roughly $30 billion in Intel and about $21 billion in SpaceX.
The bet nobody clapped for
Here is the part I like. Back in September, Nvidia put about $5 billion into Intel — the aging chip giant a lot of clever people had already written off for dead. There was no confetti. Intel was the has-been; Nvidia was the darling. Buying your struggling old rival is not the sort of thing that gets you cheered at a conference.
By the end of June, with Intel trading near $140 a share, that $5 billion was worth close to $30 billion. That is roughly a $25 billion unrealized gain in under seven months.
Price is what you pay. Value is what you get. Somebody in Santa Clara understands the difference.
Now, I want to be careful. An unrealized gain is a number on a page, not money in the bank, and any stock that rises fivefold in half a year can give some of it back in an afternoon. But the logic underneath the trade is sound, and that is what matters. Nvidia and Intel agreed to build chips together for years to come. When you invest in a partner whose success you are actively helping to manufacture, you are not gambling. You are tending your own garden.
The rocket money
The $21 billion SpaceX position rhymes with the same tune. Elon Musk has said SpaceX will run on Nvidia silicon and nobody else's. So Nvidia owns a piece of a customer that is contractually devoted to buying what Nvidia sells. That is not a lottery ticket. That is a man buying the diner where he already eats breakfast every morning.
I have no crystal ball on Intel's turnaround or on rockets. Plenty can go wrong. But I will say this: while the rest of this business was busy paying the highest prices in history for the flashiest names, the biggest name of all was quietly buying the unloved and the useful. That is how fortunes are actually made — not in the applause, but in the aisles everyone else has walked past.
The investor's habit worth keeping
The useful lesson here is not about chips. It is about second opinions. Nvidia did not follow the crowd; it checked the math itself and reached its own verdict. You can do the same with any question worth answering.
That is the whole idea behind Gangsta AI: put your question to more than 30 leading models at once — ChatGPT, Claude, Gemini, Grok — and get one cross-checked answer instead of a single confident guess. Before you buy the consensus, see what the whole room actually thinks. It is, as I have long believed, cheaper than being wrong.
Sources
- Fortune — Nvidia has $21 billion SpaceX stake, $30 billion in Intel shares
- Yahoo Finance — Nvidia turns $5B Intel bet into $30B windfall; new $21B SpaceX stake revealed
- TechTimes — Nvidia discloses equity stakes in SpaceX and Intel, both exclusive chip buyers
- Hero photo: Jensen Huang (2025), The White House — public domain, via Wikimedia Commons
Originally published on Gangsta AI News. Gangsta AI lets you compare 30+ AI models side by side on a single prompt — free, no login. Try the AI model comparison tool.
Top comments (0)