Skip to content
    NVDA
    Earnings call· Jan 2026(Q4 FY26)

    NVIDIA Q4 FY26 earnings call NVDA

    Feb 25, 2026 Source

    Executive summary

    NVIDIA Q4 FY26 — Record Revenue and Data Center Growth Driven by AI Inflection

    NVIDIA delivered an outstanding quarter, driven by the accelerating transition to AI and the strong demand for its Blackwell architecture across a diverse customer base. The company expects continued sequential revenue growth throughout calendar 2026, fueled by the inflection of agentic AI and the belief that compute directly translates to revenue for customers. Strategic investments and ecosystem expansion are key to sustaining leadership and addressing the growing demand for AI infrastructure.

    Highlights

    5

    • Total revenue of $68 billion, up 73% year-over-year, accelerating from Q3.

    • Data Center revenue of $62 billion, up 75% year-over-year and 22% sequentially, driven by Blackwell and Blackwell Ultra ramp.

    • Networking revenue of $11 billion, up more than 3.5x year-over-year, reaching record levels.

    • Free cash flow of $35 billion in Q4 and $97 billion in FY26, with $41 billion returned to shareholders.

    • Professional Visualization revenue crossed $1 billion for the first time, up 159% year-over-year.

    Concerns

    2

    • Gaming segment expects supply constraints to be a headwind in Q1 and beyond, despite strong end demand.

    • No data center compute revenue from China assumed in Q1 FY27 outlook, with uncertainty on import approvals.

    Guidance & targets

    11
    CategoryTargetConfidence
    Q1 FY27 Revenue
    $78 billion, plus or minus 2%
    high materiality
    High
    Q1 FY27 GAAP Gross Margin
    74.9%, plus or minus 50 basis points
    high materiality
    High
    Q1 FY27 Non-GAAP Gross Margin
    75%, plus or minus 50 basis points
    high materiality
    High
    Full-year Gross Margins
    mid-70s
    high materiality
    Medium
    Q1 FY27 GAAP Operating Expenses
    approximately $7.7 billion
    medium materiality
    High
    Q1 FY27 Non-GAAP Operating Expenses
    approximately $7.5 billion
    medium materiality
    High
    Full-year Non-GAAP Operating Expenses Growth
    low 40s on a year-over-year basis
    medium materiality
    Medium
    Full-year FY27 GAAP and Non-GAAP Tax Rates
    between 7% and 19%
    low materiality
    Medium
    Sequential Revenue Growth
    expected throughout calendar 2026, exceeding what was included in the $500 billion Blackwell and Rubin revenue opportunity
    high materiality
    High
    Supply Commitments
    inventory and supply commitments in place to address future demand, including shipments extending into calendar 2027
    medium materiality
    High
    Rubin Platform Production Shipments
    commence production shipments in the second half of the year
    high materiality
    High

    Segment performance

    4
    SegmentRevenueYoYQoQMargin
    Data Center
    Driven primarily by sustained strength in Blackwell and the Blackwell Ultra ramp. Growth on a sequential basis was a record, adding $11 billion in data center revenue. Networking was a standout, reaching record levels for both scale-up and scale-out technologies.
    Full Year FY26 Revenue: $194 billionFull Year FY26 Growth YoY: 68%Networking Revenue: $11 billionNetworking Growth YoY: >3.5xNetworking Growth QoQ: double digitsNVLink 72 scale-up switches share of revenue: ~2/3
    		$62 billion75%22%—
    Gaming
    Driven by strong Blackwell demand and improved supply. GeForce RTX is the leading platform. Supply constraints are expected to be a headwind in Q1 and beyond.
    		$3.7 billion47%——
    Professional Visualization
    Crossed the $1 billion mark for the first time. Launched the RTX PRO 5000 Blackwell workstation.
    		$1.3 billion159%74%—
    Automotive
    Driven by robust demand for self-driving solutions. Introduced Alpamayo. Physical AI is a significant contributor to revenue.
    Physical AI contribution FY26: >$6 billion
    		$604 million6%——

    Operational metrics

    21
    Data Center Business Scale
    13x
    since FY23

    Scale of data center business since the emergence of ChatGPT in fiscal 2023.

    Annual R&D Budget
    approaching $20 billion
    annual

    Fueled by this budget, NVIDIA aims to deliver X factor leaps in performance per watt.

    GB300 NVL72 Performance per Watt
    up to 50xvs Hopper
    Q4 FY26

    Reinforces NVIDIA's inference leadership, as declared by SemiAnalysis.

    GB300 NVL72 Cost per Token
    35x lowervs Hopper
    Q4 FY26

    Reinforces NVIDIA's inference leadership, leading to lowest cost per token.

    CUDA Software Performance Improvement (GB200 NVL72)
    up to 5x better
    within 4 months

    Continuous optimization of CUDA software on GB200 NVL72.

    Blackwell Infrastructure Deployed
    nearly 9 gigawatts
    Q4 FY26

    Deployed and consumed by major cloud service providers, hyperscalers, AI model makers, and enterprises.

    Top 5 Cloud Providers & Hyperscalers CapEx (2026)
    approaching $700 billionup nearly $120 billion since start of year
    2026

    These customers collectively account for a little over 50% of NVIDIA's data center revenue.

    Sovereign AI Business Revenue
    over $30 billiontripled year-over-year
    FY26

    Expected to grow at least in line with the AI infrastructure market long-term.

    GAAP Gross Margin
    75%increased sequentially
    Q4 FY26

    Increased as Blackwell continued to ramp.

    Non-GAAP Gross Margin
    75.2%increased sequentially
    Q4 FY26

    Increased as Blackwell continued to ramp.

    GAAP Operating Expenses Growth
    16%sequentially
    Q4 FY26

    Related to new product introductions and compute and infrastructure costs.

    Non-GAAP Operating Expenses Growth
    21%year-over-year
    Q4 FY26

    Related to new product introductions and compute and infrastructure costs.

    Non-GAAP Effective Tax Rate
    15.4%below outlook
    Q4 FY26

    Primarily due to the impact of a one-time tax benefit.

    Inventory Growth
    8%quarter-over-quarter
    Q4 FY26

    Strategically secured inventory and capacity to meet demand beyond the next several quarters.

    Capital Returned to Shareholders
    $41 billion43% of free cash flow
    FY26

    Committed to returning capital while investing in technology and ecosystem.

    Stock-Based Compensation Expense
    $1.9 billion
    Q1 FY27 outlook

    Included in non-GAAP operating expenses starting Q1 FY27; foundational component of compensation.

    Meta GEM Model Impact (Facebook)
    3.5% increase
    Q4 FY26

    Advancements in Meta's GEM model drove meaningful revenue growth.

    Meta GEM Model Impact (Instagram)
    >1% gain
    Q4 FY26

    Advancements in Meta's GEM model drove meaningful revenue growth.

    Robotaxi Fleet Scaling
    millionsfrom thousands in 2025
    next decade

    Expected to create a market generating hundreds of billions of dollars of revenue.

    AI Models on Hugging Face
    1.5 million
    Q4 FY26

    All run on NVIDIA CUDA, representing a significant open-source ecosystem.

    Anthropic Revenue Growth
    10x
    in a year

    Company is severely capacity constrained due to incredible demand and token generation rate.

    Industry KPIs

    10
    MetricValueDetails
    Lead timestightness in supply
    Ai data center revenue$62 billionUSD
    Market share commentaryInference King
    Fab capacity utilizationnearly 9 gigawattsGW
    Wfe industry spend outlookapproaching $700 billionUSD
    Design wins socket pipelineRubin samples shipped
    Inventory channel inventory8%%
    Node platform ramp scheduleRubin platform production
    End market segment revenue mix$62 billion (Data Center); $3.7 billion (Gaming); $1.3 billion (Professional Visualization); $604 million (Automotive)USD
    Strategic supply agreements customer prepaymentsinventory and supply commitments in place

    Product announcements

    4
    ProductTypeDetails
    Rubin Platformlaunch
    RTX PRO 5000 Blackwell workstationlaunch
    Alpamayolaunch
    NVIDIA Cosmos and Isaac Grouplaunch

    Deals & partnerships

    8
    OpenAIpartnership

    OpenAI's GPT-5.3-Codex trained with and inferencing on Grace Blackwell NVLink 72 systems. NVIDIA is working towards a partnership agreement.

    Metapartnership

    Meta is deploying millions of Blackwells and Rubin GPUs, NVIDIA CPUs, and Spectrum X Ethernet for training and inference.

    Anthropicinvestment and partnership$10 billion

    NVIDIA made a $10 billion investment in Anthropic. Anthropic will train and inference on Grace Blackwell and Vera Rubin systems.

    Groklicensing agreement and acquisition (talent)

    Nonexclusive licensing agreement for Grok's low latency inference technology. Team of brilliant engineers joined NVIDIA. NVIDIA will extend its architecture with Grok's innovations.

    AWSpartnership

    NVIDIA announced that it will enable AWS with NVLink to integrate with their custom silicon.

    Dassault Systemespartnership

    Expanding partnership to bring NVIDIA AI infrastructure, Omniverse digital twins, World Models, and CUDA-X libraries to researchers, designers, and engineers.

    Siemenspartnership

    Expanding partnership to bring NVIDIA AI infrastructure, Omniverse digital twins, World Models, and CUDA-X libraries to researchers, designers, and engineers.

    Synopsyspartnership

    Expanding partnership to bring NVIDIA AI infrastructure, Omniverse digital twins, World Models, and CUDA-X libraries to researchers, designers, and engineers.

    Risks & headwinds

    3
    Gaming supply constraintsQ1 FY27 and beyond

    expected to be a headwind

    China data center compute revenue uncertaintyQ1 FY27 and potentially beyond

    not assuming any revenue from China in Q1 FY27 outlook; unknown whether any imports will be allowed

    Mitigation: Continue to engage with U.S. and China government and advocate for America's ability to compete.

    Competition in Chinaover the long term

    competitors bolstered by recent IPOs are making progress and have the potential to disrupt the structure of the global AI industry

    Mitigation: America must engage every developer and be the platform of choice for every commercial business, including those in China.

    Q&A highlights

    8

    Given the high cloud CapEx approaching $700 billion and potential cash flow compression, how confident is NVIDIA in customers' ability to sustain CapEx growth, and can NVIDIA grow if CapEx doesn't?

    Jensen Huang expressed confidence in customers' cash flow growth, stating that in the new AI world, 'compute is revenues.' The inflection of agentic AI and profitable tokens drives extreme urgency for compute, shifting classical computing CapEx to AI, which requires significantly more capacity.

    “In this new world of AI, compute is revenues. Without compute, there's no way to generate tokens. Without tokens, there's no way to grow revenues.”

    asked by Vivek Arya · answered by Jen-Hsun Huang

    2 min read6 chapters

    Detailed Narrative

    01

    AI Infrastructure and Ecosystem Expansion

    NVIDIA is positioning itself as an AI infrastructure company, offering computing platforms across various aspects from GPUs to DPUs and networking. The company's strategy involves investing in its ecosystem across the entire stack, including AI models, networking, and DPU, to ensure all AI ecosystems are built on top of NVIDIA's platform. This approach aims to expand and deepen its reach in areas like language AI, physical AI, physics, biology, and robotics for manufacturing, leveraging the versatility of CUDA.

    02

    Agentic AI Inflection Point

    The company highlights the recent inflection of agentic AI, with systems like OpenAI's GPT-5.3-Codex and Anthropic's Claude Cowork driving exponential growth in token generation and compute demand. This shift is translating directly into revenue for cloud service providers and model makers, as compute capacity is now directly linked to revenue generation. The rapid adoption and profitability of tokens are creating extreme urgency to scale compute, making inference performance critical.

    03

    Networking Leadership and Innovation

    NVIDIA's networking business, including NVLink, Spectrum-X Ethernet, and InfiniBand, has seen significant growth, with Q4 revenue up over 3.5x year-over-year. The company asserts itself as potentially the largest Ethernet networking company globally, driven by its AI infrastructure design and the ability to offer superior performance per dollar and watt. This enhanced networking capability is crucial for maximizing AI factory utilization and revenue, especially in large-scale data center deployments.

    04

    Strategic Capital Allocation

    NVIDIA emphasizes a strategic and disciplined approach to capital allocation, prioritizing investments in technology and its ecosystem to cultivate market development and drive long-term growth. While committed to returning capital to shareholders through buybacks and dividends, the primary focus remains on supporting the ecosystem and ensuring supply for advanced architectures. This includes securing inventory and capacity to meet demand extending into calendar 2027.

    05

    Future of Computing and AI Factories

    Jensen Huang articulates a vision where AI is the future of computing, requiring significantly more computation than classical methods, with computing demand being 1,000 times higher. He posits that every company will depend on AI and produce 'tokens,' leading to the concept of 'AI factories' that generate revenue. This new industrial revolution, driven by agentic AI and eventually physical AI, necessitates continuous investment in compute capacity, with NVIDIA's architecture designed to maximize performance per watt and dollar.

    06

    China Market Dynamics

    While small amounts of H200 products for China-based customers were approved by the U.S. government, NVIDIA has not yet generated revenue from these and remains uncertain about future import allowances. The company acknowledges the progress of Chinese competitors, bolstered by IPOs, and their potential to disrupt the global AI industry. NVIDIA continues to engage with both U.S. and China governments, advocating for America's ability to compete globally.

    AI-generated summary of the company’s earnings call. Not investment advice.