NVIDIA NASDAQ: NVDA reported record results for its fourth quarter of fiscal 2026, driven by continued expansion in data center demand and a broadening customer mix that now includes cloud providers, hyperscalers, AI model makers, enterprises, and sovereign nations. On the company’s earnings call, CFO Colette Kress said total revenue rose to $68 billion, up 73% year-over-year, while the company also posted record operating income and free cash flow.
Record data center revenue powered by Blackwell ramp
Kress said the quarter’s sequential growth was a record, with NVIDIA adding $11 billion in data center revenue versus the prior quarter. Fourth-quarter data center revenue totaled $62 billion, up 75% year-over-year and 22% sequentially, driven primarily by “sustained strength in Blackwell and the Blackwell Ultra ramp.” She added that even Hopper products and “much of the six-year-old Ampere-based products are sold out in the cloud” due to high demand for NVIDIA infrastructure.
On a full-year basis, data center generated $194 billion of revenue, up 68% year-over-year. Kress said NVIDIA has scaled its data center business by nearly 13x since fiscal 2023, which she characterized as the period following the emergence of ChatGPT.
NVIDIA also emphasized performance-per-watt improvements as a key purchasing driver in power-constrained data centers. Kress cited third-party assessments, noting SemiAnalysis called NVIDIA “Inference King,” and pointed to results from InferenceX indicating that GB300 NVL72 achieved up to 50x performance per watt and 35x lower cost per token compared with Hopper. She also said ongoing CUDA optimization delivered up to five times better performance on GB200 NVL72 within four months.
Networking becomes a larger pillar of the AI infrastructure business
Networking was a standout in the quarter, generating $11 billion in revenue, up more than 3.5x year-over-year, according to Kress. She said both “scale up and scale out technologies” grew double digits sequentially, supported by adoption of NVLink, Spectrum-X Ethernet, and InfiniBand. She attributed year-over-year growth primarily to NVLink 72 scale-up switches, adding that Grace Blackwell systems represented roughly two-thirds of data center revenue in the quarter.
Jensen Huang, NVIDIA’s CEO, described networking as an extension of the company’s rack-scale AI infrastructure approach, saying NVIDIA “doesn’t ship nodes of computers, we ship racks of computers.” He also said NVLink “turbocharged” the networking business due to the amount of switching required per rack. Huang called Spectrum-X Ethernet “a home run,” while also emphasizing continued support for InfiniBand for customers prioritizing ultra-low latency and scale-up capabilities.
Demand drivers: agentic AI, hyperscale upgrades, and sovereign AI
Management repeatedly tied demand to the rise of agentic AI and expanding inference workloads. Kress said “agentic and physical AI applications” built on increasingly capable models are beginning to drive financial performance. She also pointed to evidence of ROI as hyperscalers upgrade major traditional workloads—such as search, ad generation, and recommender systems—to generative AI.
Kress cited Meta as an example, saying advancements in its GEM model drove a 3.5 increase in ad clicks on Facebook and a gain of more than 1% in conversations on Instagram, translating into meaningful revenue growth. She added that analyst expectations for 2026 CapEx among the top five cloud providers and hyperscalers—customers that collectively represent “a little over 50%” of NVIDIA’s data center revenue—have increased by nearly $120 billion since the start of the year and are approaching $700 billion.
In the Q&A, Huang said he is confident hyperscalers can continue investing, arguing that “compute is revenues” in the token-driven AI economy. He said agentic AI has reached an inflection point and that “we’re generating profitable tokens that are productive for customers and profitable for the cloud service providers.”
NVIDIA also discussed sovereign AI as a growing contributor. Kress said the company’s sovereign AI business more than tripled year-over-year in fiscal 2026 to over $30 billion, driven mainly by customers in Canada, France, the Netherlands, Singapore, and the U.K.
On China, Kress said small amounts of H200 products for China-based customers were approved by the U.S. government, but NVIDIA has not generated any revenue and does not know whether imports will be allowed. She also warned that China-based competitors “have the potential to disrupt the structure of the global AI industry over the long term.”
Rubin platform, partner momentum, and Groq licensing
NVIDIA outlined its next platform transition, highlighting the Rubin platform unveiled at CES. Kress said Rubin includes six chips—Vera CPU, Rubin GPU, NVLink 6 Switch, ConnectX-9 SuperNIC, BlueField-4 DPU, and Spectrum-6 Ethernet Switch—and is designed to train MoE models with one-quarter the number of GPUs and reduce inference token cost by up to 10x compared to Blackwell. NVIDIA shipped its first Vera Rubin samples to customers earlier in the week and said it remains on track for production shipments in the second half of the year.
Huang also discussed expanding partnerships with “frontier model makers,” including OpenAI, Meta, Anthropic, and xAI. He said OpenAI recently launched GPT-5.2-Codex trained and inferencing on Grace Blackwell systems, and noted NVIDIA is working toward a partnership agreement with OpenAI. He also said NVIDIA invested $10 billion in Anthropic and announced a non-exclusive licensing agreement with Groq for low-latency inference technology, adding that NVIDIA plans to extend its architecture with Groq’s innovations and expects to share more at GTC.
Margins, cash flow, capital return, and Q1 outlook
NVIDIA reported GAAP gross margin of 75% and non-GAAP gross margin of 75.2%, which Kress said increased sequentially as Blackwell ramped. She said GAAP operating expenses rose 16% sequentially, reflecting new product introductions and compute and infrastructure costs. The company also generated $35 billion in free cash flow in Q4 and $97 billion for the full year, returning $41 billion (or 43% of free cash flow) via share repurchases and dividends.
For the first quarter of fiscal 2027, Kress guided total revenue to $78 billion plus or minus 2%, with most growth expected from data center. She said the company is not assuming any data center compute revenue from China in its outlook. NVIDIA expects GAAP and non-GAAP gross margins of 74.9% and 75%, respectively, plus or minus 50 basis points, and GAAP and non-GAAP operating expenses of approximately $7.7 billion and $7.5 billion, including $1.9 billion of stock-based compensation.
Kress also said starting this quarter NVIDIA will include stock-based compensation expense in its non-GAAP results. In gaming, she reported revenue of $3.7 billion, up 47% year-over-year, but warned that supply constraints are expected to be a headwind in Q1 and beyond. Professional visualization reached $1.3 billion in revenue, up 159% year-over-year, while automotive revenue was $604 million, up 6% year-over-year.
About NVIDIA NASDAQ: NVDA
NVIDIA Corporation, founded in 1993 and headquartered in Santa Clara, California, is a global technology company that designs and develops graphics processing units (GPUs) and system-on-chip (SoC) technologies. Co-founded by Jensen Huang, who serves as president and chief executive officer, along with Chris Malachowsky and Curtis Priem, NVIDIA has grown from a graphics-focused chipmaker into a broad provider of accelerated computing hardware and software for multiple industries.
The company's product portfolio spans discrete GPUs for gaming and professional visualization (marketed under the GeForce and NVIDIA RTX lines), high-performance data center accelerators used for AI training and inference (including widely adopted platforms such as the A100 and H100 series), and Tegra SoCs for automotive and edge applications.
Recommended Stories
This instant news alert was generated by narrative science technology and financial data from MarketBeat in order to provide readers with the fastest reporting and unbiased coverage. Please send any questions or comments about this story to contact@marketbeat.com.
Before you consider NVIDIA, you'll want to hear this.
MarketBeat keeps track of Wall Street's top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis. MarketBeat has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on... and NVIDIA wasn't on the list.
While NVIDIA currently has a Buy rating among analysts, top-rated analysts believe these five stocks are better buys.
View The Five Stocks Here
Click the link to see MarketBeat's list of ten stocks that are set to soar in 2026, despite the threat of tariffs and other economic uncertainty. These ten stocks are incredibly resilient and are likely to thrive in any economic environment.
Get This Free Report
