Sam Altman Swears Eternal Loyalty to NVIDIA (While Building His Own Factories)

Sam Altman (Sam Altman) suddenly decided to turn a new leaf, or rather, dispel the thick fog of rumors that relations between OpenAI and NVIDIA had hit a snag. In a recent post, he literally showered compliments, calling Jensen Huang's products the best AI chips in the world. Against the backdrop of endless talk that OpenAI is allegedly unhappy with its dependence on a single supplier and is preparing an "NVIDIA killer" under the codename Tigris, this statement looks like a masterclass in corporate diplomacy.

Why is this happening now? For the past year, the industry has been whispering that Altman chafes under NVIDIA's monopoly. When training next-generation models that will require computing power an order of magnitude higher than current levels is on your agenda, you need hundreds of thousands of accelerators.

Waiting for H100 shipments has become the main bottleneck for the entire industry, and OpenAI is no exception. Logically, Altman began looking for ways out, flying around the world collecting trillions of dollars for building his own factories. But the reality is that building your own infrastructure from scratch is not just about hiring brilliant engineers—it's years of testing, mistakes, and billions in burned capital.

Altman directly stated that OpenAI plans to remain a "super-customer" of NVIDIA for a very long time. This sounds like a sober admission of surrender to current reality: at this point, adequate alternatives simply don't exist. Even if custom solutions from Google or Amazon show decent results in narrow tasks, NVIDIA holds the market thanks to its CUDA ecosystem.

This is the very software layer that has become the de-facto standard for any developer. Switching to different hardware today means rewriting enormous amounts of code and losing ground in the race against Anthropic and Google, something OpenAI cannot afford. It's interesting to watch this game of thrones in the silicon world.

On one hand, we hear news about OpenAI's partnership with Broadcom and reserving capacity at TSMC factories. On the other—Altman publicly confesses his love for the current monopolist. This is a classic diversification strategy in the spirit of "keep your friends close and your enemies closer."

While OpenAI builds its backup airfield, it's vital for them that Huang continues to ship new Blackwell chips in priority order. Any hint of conflict could result in OpenAI's queue for scarce hardware suddenly becoming longer. In the end, we have a situation where Altman's words are not just politeness, but a calculated business move.

NVIDIA benefits from having such an ambassador who confirms their dominance at the highest level, and OpenAI benefits from maintaining its status as a favorite customer. While the whole world waits for a real competitor to NVIDIA to appear, the duo of Altman and Huang continues to dictate the rules of the game, even if behind closed doors each of them is already preparing a plan for world domination alone. In this industry, love lasts only as long as a more efficient technology process doesn't appear.

The bottom line: OpenAI's own chips are a dream of the future, while NVIDIA is a harsh necessity of the present. Will Altman ever truly kick his "habit" of Jensen Huang's dependence, or will this dependency prove fatal for the company?