NVIDIA commits $40B to AI partners

Nvidia is no longer just selling chips. It is funding the whole AI buildout around those chips — data centers, fiber, optics, and even some of the companies that will end up buying Nvidia systems. That is the real story behind the new reporting that Nvidia has already topped $40 billion in AI equity commitments in 2026. ### Where does the $40 billion number come from? The headline comes from CNBC’s tally of Nvidia’s 2026 dealmaking. The count includes a string of public-equity agreements announced this month plus a much larger OpenAI commitment that has been hanging over the market since September 22, 2025, when Nvidia and OpenAI said Nvidia intended to invest up to $100 billion progressively as OpenAI deployed at least 10 gigawatts of Nvidia-based AI infrastructure. (cnbc.com) Recent coverage treats roughly $30 billion of that as the anchor piece of the 2026 total. ### Why is OpenAI the center of it? Because OpenAI is not just another customer. It is one of the biggest future buyers of AI compute on the planet. The 2025 partnership tied Nvidia’s capital directly to OpenAI’s infrastructure rollout, with the first gigawatt targeted for the second half of 2026 on the Vera Rubin platform. Basically, Nvidia is helping finance demand for its own next wave of systems while locking in a flagship user for them. (cnbc.com) ### What else did Nvidia just fund? Two fresh examples landed this week. On May 7, Nvidia and IREN announced a partnership aimed at deploying up to 5 gigawatts of AI infrastructure, and IREN gave Nvidia the right to invest up to $2.1 billion through a five-year stock purchase right. Around the same time, Nvidia struck a Corning deal that gives it the right to invest up to $3.2 billion while Corning expands U.S. optical manufacturing for AI infrastructure. (investor.nvidia.com) ### Why Corning and IREN? Because AI factories are constrained by more than GPUs. They need land, power, cooling, buildings, fiber, optical interconnects, and operators that can actually stand them up fast. IREN gives Nvidia a route into powered data-center capacity. Corning gives it tighter control over optical components that move data inside and between giant clusters. The catch is that these are classic bottleneck businesses — boring until they are suddenly the thing holding back the whole build. (nvidianews.nvidia.com) ### Is this normal vendor behavior? Not at this scale. Nvidia has the cash to do it — fiscal 2026 revenue hit $215.9 billion, and the company generated huge profits and cash flow. So instead of waiting for the ecosystem to catch up, Nvidia is underwriting parts of it directly. That makes the company look less like a component supplier and more like the capital allocator for the AI stack. (nvidianews.nvidia.com) ### Why are some people uneasy? Because the model can look circular. Nvidia invests in companies, those companies build AI capacity, and that capacity often gets filled with Nvidia hardware. CNBC noted that critics compare parts of this to vendor financing from earlier tech bubbles. Wedbush analyst Matthew Bryson called it part of a “circular investment theme,” which is Wall Street shorthand for growth that may be reinforcing itself through financing as much as through end demand. (nvidianews.nvidia.com) ### So what changed this week? The new thing is visibility. Nvidia’s IREN and Corning announcements made the pattern hard to miss, and the CNBC tally turned a bunch of separate deals into one clear number — more than $40 billion in just the opening months of 2026. Once you see it that way, Nvidia’s strategy looks less accidental and more like a deliberate attempt to own the choke points around AI infrastructure. (cnbc.com) ### Bottom line? Nvidia is using its balance sheet to make sure the AI boom has enough power, fiber, buildings, and customers to keep buying Nvidia gear. That is smart — but it also means the company is now shaping the market it sells into, not just serving it. (cnbc.com)