Nvidia and OpenAI: Jensen Huang pulls out the checkbook, but not for 100 billion

Let's be honest: the $100 billion figure that's been making headlines over the past few days looked like pure hallucination from an overheated market. Even for Nvidia, which is literally printing money on the sales of its H100s, such a sum within the scope of a single deal looks absurd. Nevertheless, where there's smoke, there's fire. Jensen Huang set the record straight this week, confirming that Nvidia is indeed planning to become one of OpenAI's largest investors in the current funding round. In doing so, he skillfully dodged confirming specific amounts, but decisively refuted rumors that the company was allegedly "doubting" the wisdom of this step.

To understand why this alliance is so important, you need to go back to 2016. Back then, Jensen Huang personally delivered the first DGX-1 supercomputer to OpenAI's office, where it was received by an Elon Musk and Sam Altman who hadn't yet fallen out. Since then, these two companies have been connected by an umbilical cord: OpenAI creates demand for computational power that no one but Nvidia can satisfy.

Now, when OpenAI is valued at $150 billion, Nvidia simply can't afford to stay on the sidelines. This is not a question of profit, but of control over the ecosystem. When your main client is simultaneously your main testing ground, you want to have a seat on the board of directors or at least a very weighty voice in the room.

The intrigue surrounding Nvidia's "doubts" did not arise from nowhere. The market feared that Jensen wouldn't like OpenAI's excessive rapprochement with Microsoft or Sam Altman's attempts to launch his own chip manufacturing project. But Huang is a pragmatist. He understands that even if OpenAI starts designing its own processors, it will take years to reach actual production capable of competing with Blackwell architecture. Right now, it's more important for Nvidia to keep OpenAI in its orbit so that CUDA architecture remains the de facto standard for training the world's largest models. Investments here work like an insurance policy: Nvidia puts in money that OpenAI immediately spends back on it, buying new batches of graphics processors.

It's also interesting how this deal changes the landscape of venture capital in Silicon Valley. We're witnessing the birth of a new "investor-supplier" model. Nvidia, through its NVentures division, has already invested in dozens of AI startups, essentially creating a closed loop of consumption of its product. OpenAI is just the biggest fish in this pond. If the deal closes on expected terms, Nvidia will find itself in the same boat as Thrive Capital, Apple, and Microsoft. This transforms OpenAI into a kind of "holy grail" of the industry, in whose success all the giants are directly invested, even if in other spheres they're ready to tear each other's throats out.

What does this mean for us? Most likely, we'll see even tighter integration of OpenAI's software and Nvidia's hardware. This is bad news for competitors like AMD or Intel, for whom it will be even harder to knock this tandem off its feet. Jensen Huang once again proved that he's not just selling "shovels" during a gold rush; he's buying the mines themselves. And although the $100 billion turned out to be a newspaper hoax, the actual amount of investment will still be a clear signal: the era of Nvidia's dominance is only just beginning, and they're willing to pay to make sure the rules of the game are dictated by them.

The bottom line: Nvidia is finally transitioning from vendor status to that of the main shareholder of the AI future. Will anyone even be able to compete with a coalition where hardware, software, and clouds are merged into a single financial fist?