NVDA

Q3 2026

Date: 2025-11-19Rank: #10Forward Promise: very_bullish

NVIDIA delivered a record quarter with revenue of $57 billion, up 62% year-over-year and 22% sequentially, driven by a 66% YoY increase in Data Center revenue to $51 billion. The Blackwell architecture ramp is the primary growth driver, with the GB300 now contributing two-thirds of Blackwell revenue, while networking revenue more than doubled. Management announced a strategic partnership with OpenAI to build 10+ gigawatts of data centers and a new deal with Anthropic to optimize their models on CUDA. Looking forward, the company guided for Q4 revenue of $65B (+/- 2%), implying 14% sequential growth, and stated visibility to $500 billion in Blackwell and Rubin revenue through the end of calendar 2026.

Bullishness Score

95.98

μ Mean

101.12

σ Uncertainty

1.71

Forward Promise

8.5

Management Tone

Management exuded high confidence and a sense of inevitability regarding their market leadership. Jensen Huang was particularly energetic and assertive, framing the current environment not as a bubble but as three simultaneous, massive platform shifts that NVIDIA is uniquely positioned to dominate. The tone shifted from purely financial celebration in prepared remarks to a defensive yet spirited defense of the AI investment cycle and NVIDIA's architectural superiority during Q&A.

Confidence: HIGH — Management consistently used definitive language regarding demand visibility and product execution, dismissing concerns about bubbles or ASIC competition with specific technical and economic rebuttals.

Strategic Signals

NVIDIA is aggressively expanding its 'AI Factory' concept, moving beyond selling chips to building massive infrastructure partnerships. The announcement of a strategic partnership with OpenAI to develop 10+ gigawatts of data centers signals a deepening relationship that secures massive future demand. Similarly, the new partnership with Anthropic, bringing them onto the NVIDIA platform for the first time, expands NVIDIA's reach in the enterprise AI sector. These moves indicate a strategy to lock in the world's leading AI model builders into the NVIDIA ecosystem.

The company is executing a rapid product cadence to maintain performance leadership. The GB300 has already overtaken the GB200 in revenue contribution just one quarter after the latter's launch, demonstrating an aggressive upgrade cycle. Management confirmed that the Rubin platform is on track for 2026, with silicon already received from supply chain partners. This annual 'x-factor' performance improvement is designed to render competitors' products obsolete before they reach scale.

NVIDIA is leveraging its full-stack advantage to defend against ASIC competition. During the call, Huang outlined five specific moats: accelerating every phase of computing (CPU to GPU, GenAI, Agentic), excelling at every phase of AI (pre-training, post-training, inference), running every model type, being available in every cloud, and providing diverse offtake. This comprehensive positioning is intended to convince customers that general-purpose GPUs are more versatile and economically viable than custom accelerators.

Networking is emerging as a major growth driver, with revenue up 162% year-over-year. Management highlighted that the majority of AI deployments now include NVIDIA switches, with Ethernet attach rates matching InfiniBand. The introduction of Spectrum XGS for 'gigascale' AI factories and the integration of NVLink IP into partners like Arm and Intel suggest NVIDIA is successfully building a proprietary networking moat that complements its compute dominance.

Key Metrics

Total Revenue$57.0B+62% YoY / +22% QoQ

Data Center Revenue$51.0B+66% YoY

Networking Revenue$8.2B+162% YoY

Gaming Revenue$4.3B+30% YoY

GAAP Gross Margin73.4%+ sequentially

Non-GAAP Gross Margin73.6%+ sequentially

Q4 Revenue Guidance$65.0B (+/- 2%)+14% QoQ at midpoint

China Data Center Revenue$2.0BSignificant decline YoY

Guidance

Q4 Revenue: $65 billion +/- 2%

Q4 Gross Margin: 74.875% +/- 50 bps

FY27 Gross Margin: Mid-70s (working to hold)

Blackwell/Rubin Visibility: $500B through end of calendar 2026

Bottom Line

Price Target: N/A

NVIDIA remains the undisputed leader in AI infrastructure, and Q3 proved that the growth trajectory is steepening rather than slowing. The seamless transition to the GB300 architecture, which is already driving two-thirds of Blackwell revenue, demonstrates execution excellence. The strategic partnerships with OpenAI and Anthropic effectively lock in the world's most sophisticated AI consumers into the NVIDIA ecosystem for years to come. While margin pressure in FY27 and the loss of the China market are valid concerns, the visibility to $500B in revenue through 2026 provides a clear floor for growth. The company is successfully expanding its moat by moving into networking and software, making it increasingly difficult for competitors or customers to switch away. The current valuation is supported by the specificity and confidence of management's forward guidance.

Forward Promise

Overall: 8.5 · very_bullish

Unapologetically bullish: 'The number will grow' on $500B visibility.

Management demonstrated exceptional control over the narrative, providing deep technical and economic context to support the view that the AI boom is durable. The specific reaffirmation of the $500B roadmap and the seamless GB300 transition signal strong near-term momentum.

Red Flags

Gross margin sustainability is a primary concern for investors heading into fiscal 2027. Management admitted that input costs are rising and guided that they are 'working to hold' gross margins in the mid-seventies, implying potential pressure. While they cited cost improvements and mix as mitigants, the shift to the more complex Blackwell and Rubin architectures involves higher packaging and memory costs (HBM), which could compress profitability if pricing power wanes.

The geopolitical situation in China remains a significant headwind. Management noted that 'sizable purchase orders never materialized' due to geopolitical issues and increasing competition, resulting in only $2 billion in revenue from a region that was previously a massive growth driver. While the guidance assumes no data center compute revenue from China, the loss of this market or the emergence of strong domestic competitors (like Huawei) creates a persistent risk to the growth narrative.

The sheer scale of the $500 billion visibility figure raises execution risk. While management claims visibility to this amount through 2026, the company is dependent on a complex, global supply chain to deliver. Any hiccup in the yield of the new Blackwell or Rubin chips, or shortages in HBM memory or advanced packaging capacity, could derail these aggressive shipment targets. The 63% sequential increase in supply commitments indicates a massive ramp that leaves little room for error.

Customer concentration risk remains elevated despite the expansion into sovereigns and enterprises. The narrative is still heavily reliant on the 'hyperscalers' and a few foundation model builders (OpenAI, Anthropic, xAI) for the bulk of the projected growth. If these few players reduce their CapEx intensity due to ROI concerns or internal delays, NVIDIA's revenue could decelerate rapidly, as evidenced by the analyst questions regarding funding and ROI.

Macro Insights

AI CapEx: Management estimates hyperscaler CapEx in 2026 at ~$600B, up $200B from the start of the year, indicating no slowdown in infrastructure spending.

Power Infrastructure: Power availability is a key bottleneck, but NVIDIA is mitigating this through performance-per-watt efficiency gains and partnerships for 'behind the meter' power solutions.

Geopolitics: Geopolitical tensions and competition in China prevented sizable orders from materializing, effectively removing this market from the growth story for now.

Enterprise Adoption: Agentic AI is driving a new wave of enterprise adoption, with companies like Salesforce, Unilever, and RBC reporting significant productivity gains, justifying further investment.