Nvidia Writes Checks for the AI Boom It Expects to Supply

Nvidia commits $2B to a Brookfield AI fund, expands data centers and software ties, as Q2 revenue hits $96.2B with no China data center sales assumed.

Nvidia is no longer content to sit at the end of the AI supply chain and collect orders. Within a matter of weeks, the chipmaker has moved on two fronts — committing capital to the infrastructure its chips will populate, and embedding its software deep enough into customer operations that switching away becomes a strategic headache rather than a procurement decision.

The most striking example is a reported $2 billion commitment to an AI fund run by Brookfield, the investment firm, according to media accounts citing Bloomberg. The logic is circular in a way that favors Nvidia: a meaningful slice of that money is expected to flow back through the construction of massive server farms and the purchase of new accelerator chips. Rather than waiting for hyperscalers or governments to find the cash, Jensen Huang's company is priming the pump itself.

Building the Ground It Will Later Sell

That financial muscle is being matched by a widening geographic footprint. In Australia, Nvidia plans to expand data center capacity through partnerships, a move Reuters reports would more than double utilization there. In Southeast Asia, the company secured a credit facility of up to $3.1 billion for Zankore to finance cloud infrastructure built on Nvidia GPUs in Indonesia.

Standards are part of the same play. Chip developer D-Matrix will adopt NVLink Fusion so its processors can be integrated into Nvidia systems from 2027 onward — another instance of Nvidia's interfaces becoming the industry's default plumbing.

The software layer is advancing just as aggressively. A partnership sealed on September 10 with Palantir Technologies pairs Palantir's sovereign AI software with Nvidia's own Nemotron models, initially to harden Nvidia's critical supply chains. A day earlier, the two companies had already moved to build a joint operational AI stack for enterprises and government agencies. Running the technology through its own operations first gives Nvidia both a stress test and a reference case.

Meanwhile, the Vera Rubin platform is entering production with partners including Google Cloud, Microsoft Azure and Oracle Cloud Infrastructure — deepening dependencies that extend well past the purchase of a single processor.

The Numbers Behind the Ambition

The financial backdrop is anything but modest. For the second quarter of fiscal 2027, Nvidia reported revenue of $96.2 billion and a gross margin of 75.0%. Roughly $26.0 billion flowed back to shareholders during the quarter through buybacks and cash dividends, with $99.0 billion still authorized for repurchases. A quarterly dividend of $0.25 per share is scheduled for payout on October 1.

Guidance for the third quarter stands at $108.0 billion in revenue — but with an explicit caveat: no data center revenue from China is assumed. That omission cuts both ways, showcasing operational strength while laying bare the geopolitical ceiling on growth.

A Market That Flinches, Then Shrugs

Investors have shown they are quick to react to any hint of a wobble in the AI investment cycle. On Monday, the stock fell 3.4% as warnings from leading industry figures about the risks of cutting-edge AI made the rounds, stoking fears of softening demand for expensive computing power. The dip proved short-lived. Shares closed Friday at EUR 193.10, up about 20% year to date and just 4.6% below their 52-week high.

That valuation leaves little room for error. A company priced for near-flawless execution cannot afford operational delays or regulatory setbacks, and the dual role of equipment supplier and financier adds real complexity. If AI demand cools unexpectedly, Nvidia would feel it on several levels at once.

Analysts remain largely undeterred. Piper Sandler initiated coverage on September 10 with an Overweight rating and a $300 price target. The next test comes on Tuesday, November 17, when quarterly results are due.

The strategic case, though, rests on more than any single quarter. By funding data centers worldwide, locking in standards like NVLink and stitching together global partnerships, Nvidia is constructing an industrial ecosystem that competitors will find hard to dislodge — a moat deeper than any advantage measured in raw chip performance.