Nvidia's first earnings report since Jensen Huang resigned as chief executive combined record financial results with an abrupt leadership transition, as the company posted quarterly revenue above $100 billion for the first time and forecast another sharp increase in sales. The board unanimously accepted Huang's resignation, appointed Colette Kress as interim CEO while she continues as chief financial officer, and began a formal search for a permanent replacement.
For the fiscal 2027 third quarter, Nvidia reported $115.8 billion in revenue, exceeding the $109.1 billion market estimate cited in the company's earnings materials. Revenue rose 103% from a year earlier and 20% from the previous quarter. Net income reached $65.3 billion, up 115% year over year and 9% sequentially, compared with a $60.406 billion market estimate.
The results prompted Nvidia shares to rise more than 4% in after-hours trading. The immediate market reaction reflected both the earnings beat and a fourth-quarter forecast that pointed to continued demand for AI computing systems despite mounting constraints in chips, packaging and server deployment capacity.
Fourth-quarter outlook reaches $130 billion
Nvidia projected fiscal fourth-quarter revenue of $130 billion, above the $125.2 billion market estimate cited by the company. Kress also raised Nvidia's fiscal 2028 revenue growth outlook to roughly 75%, from an earlier estimate of 70%, during the earnings call.
Data center revenue remained the dominant engine of the business, reaching $107.6 billion in the third quarter. That figure was up 110% from a year earlier and 21% from the preceding quarter, meaning the segment alone accounted for nearly all of Nvidia's quarterly sales.
Kress said cloud-industry backlog had risen to $2.5 trillion, from $2 trillion three months earlier. The $500 billion quarterly increase illustrates the scale of infrastructure commitments being made by cloud operators and AI developers, while Nvidia said demand for its products continued to exceed available supply.
The company attributed demand to hyperscalers, specialized "neocloud" providers, frontier AI laboratories, enterprises and national governments. It expects frontier model labs to represent about 15% of data center revenue this year, placing a sizable part of Nvidia's sales growth in the hands of a small group of companies developing large AI models.
Nvidia cited Anthropic's annualized revenue run rate of $81 billion as of late October as an example of the commercial scale emerging around AI model providers. It also said computing capacity supplied to hyperscalers and neocloud customers was operating at 100% utilization.
Vera Rubin platform posts rapid early sales
Nvidia said its Vera Rubin platform generated nearly $20 billion in revenue in its first quarter of mass-production shipments, making it the company's fastest platform ramp by its own measure. Vera Rubin NVL72 began mass production and shipments in August, Nvidia said, and hyperscale customers moved systems from delivery into production in around one month.
That deployment time was about half the period required for Grace Blackwell systems, according to Nvidia. Faster installation can help cloud providers convert delivered hardware into revenue-producing capacity more quickly, a meaningful factor when equipment remains in short supply.
Nvidia's network revenue surpassed $21 billion in the quarter, while the company said demand for its CPUs was also running ahead of supply. The company linked that demand to agentic AI workloads, which require more task scheduling and data processing alongside the graphical processing units used for model training and inference.
CoreWeave, Meta, Nscale and OpenAI joined Nvidia's Vera CPU customer list during the quarter, the company said. AWS, Oracle Cloud Infrastructure and SpaceX AI had previously announced adoption. Kress expects CPU revenue to more than double during fiscal 2028.
The company also described growing demand beyond large language models. Sovereign AI revenue, referring to systems bought by governments and national programs, more than doubled year over year in the quarter. Physical AI revenue, tied to robotics and AI systems operating in real-world environments, nearly doubled, with Nvidia projecting more than $25 billion in full-year physical AI revenue.
In enterprise markets, Nvidia said on-premises revenue from financial-services customers approached $5 billion over the past 12 months, more than doubling from the prior year.
Margins remain high despite supply expansion
Nvidia reported GAAP gross margin of 74.8% and non-GAAP gross margin of 74.9%, above its previous guidance of about 74%. Kress attributed the outcome partly to product mix effects from Blackwell Ultra systems.
For the fiscal fourth quarter, Nvidia forecast gross margin of 71% to 72%, which it described as a near-term low point. The company expects margins to return to a 72% to 73% range during fiscal 2028 as product transitions and supply investments progress.
GAAP diluted earnings per share came in at $2.70, while non-GAAP diluted earnings per share was $2.68. Both figures exceeded market expectations, according to Nvidia.
Supply remains a practical constraint on the revenue outlook. Kress said Nvidia is working with manufacturing partners to expand wafer, substrate and advanced-packaging capacity. During the quarter, the company added substrate agreements with Kinsus and Samsung Electro-Mechanics and an advanced-packaging agreement with SPIL. Those arrangements build on expansion plans announced in September with Amkor, KYEC and Unimicron.
The bottleneck extends beyond Nvidia's own production lines. Large AI systems require chips, networking equipment, advanced packaging, memory, power infrastructure and physical data-center space to arrive on coordinated schedules. The International Energy Agency has projected that global data centers could consume as much as 945 terawatt-hours of electricity by 2030, underscoring the power requirements attached to the hardware buildout.
Board starts CEO search after Huang resignation
Nvidia's board hired executive-search firm Heidrick & Struggles to consider internal and external CEO candidates. Harvey Jones, Nvidia's board chair, said the board does not plan to return Huang to the chief executive role.
The company also created a national security and export-controls oversight committee comprising independent directors, outside senior advisers and legal counsel. Executives did not address questions on the earnings call concerning Huang's resignation or an investigation mentioned by callers.
The leadership change leaves Kress overseeing Nvidia through a period when its revenue base, supply chain and customer commitments are expanding at unusual speed. The company's guidance indicates that demand for AI infrastructure remains far larger than the systems currently reaching deployment, but maintaining that trajectory will depend on whether Nvidia and its partners can add manufacturing, packaging, energy and data-center capacity quickly enough to meet the backlog.
Explore Nvidia-driven market opportunities and institutional flows in our latest outlook: read the full analysis.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
