Observing Elon Musk discuss Nvidia these days gives the impression that he is attempting to do two things at once. He is writing checks to Jensen Huang on one side. On the other hand, he is constructing a semiconductor plant in Texas called Terafab, which should eventually reduce those checks. Nevertheless, he tweeted that he is still a “huge admirer” of Huang on a bright Thursday in March, and the news caused Nvidia’s stock to rise. It’s difficult to ignore the weight that even one of his words still carries.
The background is important. Tesla’s first quarter was inconsistent, with earnings slightly better than anticipated and revenue softer than the Street had anticipated. The spending plan, however, is the bigger picture. In order to finance AI training, chip design, and the manufacturing foundation for Optimus and the robotaxi fleet, Musk informed analysts that capital expenditures would significantly increase. The shift is significant for a business that was primarily recognized as an automaker until recently. In fact, car sales decreased in 2025. The valuation, which trades at about 178 times forward earnings, is currently being carried by robots and self-driving software.
| Key Information | Details |
|---|---|
| Subject | Elon Musk’s recent comments on Nvidia and its stock |
| Companies Involved | Tesla (NASDAQ: TSLA), Nvidia (NASDAQ: NVDA), SpaceX, xAI |
| Nvidia CEO | Jensen Huang, co-founder, has been leading the company since 1993 |
| Tesla CEO | Elon Musk also heads SpaceX and xAI |
| Musk’s stance | “Huge admirer” of Huang; Tesla and SpaceX will keep buying Nvidia GPUs in substantial quantities |
| Tesla’s in-house effort | AI5 chip, Dojo supercomputer, and Terafab (Texas semiconductor plant with SpaceX/xAI) |
| Recent NVDA share price | Around $200, with analyst median target near $267 (NVDA quote) |
| Reported Nvidia order book | About $1 trillion in expected purchases through 2027 for Blackwell and Vera Rubin chips |
| Tesla 2026 capex guidance | Raised from $20B to over $25B |
Only when the AI wagers are successful does that kind of premium make sense. And Musk needs chips to make those wagers pay off. Many of them. Musk has publicly stated that he is not attempting to replace Nvidia’s GPUs, which are still the industry standard for training large models, but rather to supplement them with Tesla’s AI5 silicon for edge computing in vehicles and robots. Although the dual approach is now typical in big tech, it feels different coming from Musk, in part because he often speaks more forcefully about rivals. The tone softens with Huang.
Investors quickly realized that. Musk’s recent remarks were interpreted as a subdued bullish signal for Nvidia, assuring the market that one of the world’s most picky consumers of AI hardware isn’t abandoning the company. As Anthropic, Amazon, Google, and Musk’s own group promote their custom chips, there is still doubt about how long Nvidia’s lead will last. However, according to reports, Huang’s order book, which is anchored by Blackwell and the upcoming Vera Rubin generation, is currently close to a trillion dollars through 2027.
Nevertheless, there are some awkward aspects to the relationship. The peculiar situation that Nvidia might find itself in if SpaceX decides to buy the AI coding startup Cursor was recently highlighted by Bloomberg. Additionally, Musk has repeatedly requested that Nvidia reroute chips that were originally scheduled for Tesla to xAI when xAI training runs required additional processing power. One minute it’s a partner, the next it’s a rival, and then it’s a customer. Engineers at Nvidia must occasionally wonder which Musk they will hear from next outside the shiny Santa Clara headquarters.
The simplest explanation might be the correct one. In a podcast a few months ago, Musk stated that although he doesn’t actually invest in stocks, if he had to choose names “to meet a capitalistic end,” he would mention Google and Nvidia. “Nvidia is obvious at this point,” he remarked, seemingly shrugging. That’s an amazing compromise from the man constructing a competing chip stack.
For the time being, investors appear to think that Musk’s ongoing purchases are proof that the AI cycle still has room to grow. Huang’s full-stack pitch, Microsoft and Amazon following suit, and Tesla increasing its 2026 capital expenditures beyond $25 billion all point in the same direction. It remains to be seen if Terafab will ever alter that calculation. One tweet at a time, the response is currently being delayed.