Nvidia shares rebounded in after-hours trading Wednesday after initially slipping following the chipmaker's fiscal second-quarter earnings report.

The stock closed regular trading down 1.59% at $209.66 before recovering to roughly $217 after hours, about 3.5% above its closing price. Nvidia is the dominant supplier of the accelerators that power large AI models, and its results are widely treated as a barometer of spending on AI data centers across the technology industry, which is one reason its earnings reports are closely watched even when results beat expectations.

Nvidia's quarterly revenue more than doubled, rising 106% year over year to $96.2 billion and exceeding the $92.27 billion average analyst estimate reported by the Associated Press. Data Center revenue, which accounts for the overwhelming majority of Nvidia's sales, climbed 117% to $89 billion, while adjusted earnings of $2.22 per share beat the $2.09 estimate. Gross margins held at 75%.

Founder and CEO Jensen Huang attributed the quarter's growth to accelerating demand for artificial intelligence infrastructure.

"AI has reached its inflection point. It's doing useful work. Its tokens are productive and profitable. Now, compute is revenue, and demand is accelerating," Huang said in a statement. "This time last year, one lab alone was driving the buildout; today, we have a golden age of new AI labs and startups, multiple frontier labs scaling in parallel, a thriving open-model ecosystem and physical AI coming online—with strong momentum across the U.S. and around the world."

Infrastructure Commitments Reach $366 Billion

Chief Financial Officer Colette Kress said Nvidia has made $366 billion in multiyear artificial intelligence infrastructure commitments, including $279 billion for supply and capacity. The company also disclosed up to $108.5 billion in phased guarantees, which represent potential exposure rather than current losses or payments. The shift from buying chips as needed to locking in multiyear supply and capacity reflects how competition for components such as memory has intensified as AI data-center buildouts scale, and the sheer size of the figure signals how far in advance Nvidia now plans its supply chain.

"We've partnered with our extensive network of suppliers to secure the critical components needed to meet demand for the next several years," Kress said in a statement. "Our commitments increased from $119 billion last quarter to $279 billion, primarily related to the procurement of memory."

Nvidia also reported $3.5 billion in maximum gross exposure under those agreements.

In August, Nvidia agreed to support the development of SB Energy's PORTS-Pike Technology Campus in Ohio, a site expected to host Nvidia infrastructure under long-term leases to OpenAI. The arrangement is part of a broader wave of large-scale AI campus projects in the U.S., where developers are pairing dedicated power generation with data-center capacity for major AI customers.

"Each generation of NVIDIA infrastructure deployed at PORTS-Pike could represent approximately 1.5 million NVIDIA GPUs, or approximately $150 billion to $200 billion in NVIDIA revenue," she wrote. "Over 20 years, we expect the site can support multiple infrastructure upgrade cycles."

Including both arrangements, maximum gross guarantee exposure stands at $108.5 billion—a figure that represents potential exposure rather than a current loss or payment.

Nvidia Forecasts $108 Billion in Q3 Revenue

Nvidia expects third-quarter revenue of $108 billion, plus or minus 2%, which would mark another sequential step up from the just-reported quarter. It forecast gross margins of 74%, plus or minus half a percentage point, compared with 75% in the second quarter.

Shipments of Hopper data-center products to China represented less than 1% of Data Center revenue during the quarter, reflecting how U.S. export controls have curtailed Nvidia's sales of high-end accelerators into the Chinese market.

Despite the dip in the stock price, Huang said Nvidia is preparing to meet demand with its next-generation computing platform.

"The AI infrastructure buildout is at full steam," he said. "Vera Rubin, now in full production, was built to power exactly this moment."