NVIDIA’s Q2 FY2027 revenue reached $96.2 billion, a 106% increase year-over-year that underscores the sheer scale of the current infrastructure buildout. Within this total, the Data Center segment contributed $89.0 billion, representing 92% of the company’s total revenue. This performance provides the latest evidence for the compute landlord thesis, a framework explored in our previous coverage of NVIDIA’s massive exposure to OpenAI, its strategic investment in Perplexity, and the 10-gigawatt Vera Rubin deployment by SpaceX.
The ACIE Breakout and Structural Diversification
For the first time, NVIDIA has disclosed the ACIE segment—encompassing AI Cloud, Industrial, Enterprise, and Sovereign AI. This segment generated approximately $40.0 billion, accounting for 45% of total Data Center revenue. The 138% year-over-year growth in ACIE is structurally significant because it signals a shift beyond the initial hyperscale-only phase of the AI buildout. Sovereign AI, in particular, grew 35% quarter-over-quarter and has more than tripled year-over-year, indicating that national governments and regional entities are now active participants in the compute market.
This diversification is further supported by enterprise on-premise activity. The trailing twelve-month (TTM) data shows approximately $8 billion in automotive-related compute, alongside $7 billion combined from financial services, manufacturing, and healthcare. Additionally, Edge Computing contributed $7.2 billion, growing 27% year-over-year, suggesting that the demand for compute is moving closer to the point of data generation.
Financing as a Strategic Mechanism
NVIDIA is evolving from a hardware vendor into a financial architect. The company recently announced $500 billion in compute financing memorandums of understanding (MOUs) with major financial institutions, including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR. This mechanism allows NVIDIA to facilitate the massive capital expenditures required by its customers, effectively lowering the barrier to entry for large-scale infrastructure projects and ensuring a steady pipeline of demand for its hardware.
Vera Rubin in Production
The Vera Rubin platform is now in full production, serving as the physical proof point for the company’s current trajectory. Racks are currently operational at CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure, and Nebius. Beyond these cloud providers, the platform is being integrated into specialized infrastructure, such as the SpaceXAI deployment of Vera CPUs and the SB Energy partnership at the PORTS-Pike site in Ohio.
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. 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. The AI infrastructure buildout is at full steam. Vera Rubin, now in full production, was built to power exactly this moment.
Jensen Huang’s assessment highlights the transition from speculative investment to productive, revenue-generating compute. However, this rapid expansion is not without structural risks.
Risks and Margin Dynamics
While the growth trajectory remains steep, investors must account for specific headwinds. NVIDIA’s Q3 FY2027 guidance of $108 billion (plus or minus 2%) explicitly excludes China Data Center compute revenue, as noted in the company’s press release. This exclusion highlights the geopolitical constraints on the company’s addressable market. Furthermore, gross margins are expected to see slight compression, moving from 75.0% in Q2 to a projected 74.0% (plus or minus 50 basis points) in Q3. While these margins remain high by historical standards, the trend warrants monitoring as the company scales its newer platforms.
Concentration risk also remains a factor, as the hyperscale segment still accounts for approximately $49.0 billion, or 55% of Data Center revenue. Despite this, the company continues to prioritize shareholder returns, having distributed approximately $26.0 billion in the second quarter through repurchases and dividends, with $99.0 billion remaining under its current repurchase authorization.
As the compute landlord thesis enters this sixth phase, the focus shifts from whether the infrastructure will be built to how it will be financed and utilized across a broader range of sovereign and enterprise actors. The transition of compute into a primary revenue driver, supported by massive financial partnerships and the deployment of the Vera Rubin platform, suggests that the current buildout is entering a more mature, albeit more complex, operational stage.