The $2 trillion queue for data center compute is the physical manifestation of the cloud industry waiting for its next growth engine. As of Sept. 9, 2026, NVIDIA Corporation NASDAQ:NVDA has transitioned from a component vendor to the lead architect of entire AI factories. This shift moves the business beyond simple chip sales into a realm where it captures significantly more value from every gigawatt of power installed. Investors are weighing whether the rapid ramp of the Vera Rubin platform can sustain the company's valuation as rising component costs begin to squeeze margins.
Potential Bull Cases
1. The Shift to Gigawatt Scale AI Factories Reshapes the Earnings Potential
NVIDIA has evolved from selling discrete parts to offering a complete AI factory platform. In the earlier Hopper era, the company captured roughly $18 billion in revenue for every gigawatt of data center capacity. With the move to the Vera Rubin platform, that figure has soared to $40 billion per gigawatt. This expansion happens because the company now controls the entire stack, including the Vera central processing unit, networking, and software. By co-designing the hardware and the algorithms together, it captures a larger share of total data center spending. This vertical integration creates a massive structural advantage that rivals selling individual chips can't easily replicate.
2. The Vera Rubin Product Cycle Promises Unprecedented Efficiency Gains
The launch of the Vera Rubin platform marks the fastest product ramp in the company's history. These systems deliver a 30-fold increase in throughput per megawatt compared to the Grace Blackwell Ultra. For a data center operator, this efficiency translates to a 35-fold reduction in the cost to generate an AI token. Every major cloud provider and technology firm has already placed orders for these systems. Management expects Vera Rubin to account for about 20% of data center revenue in the third quarter of fiscal 2027. This rapid adoption cycle demonstrates that customers are racing to trade their older hardware for the superior economics of the latest NVIDIA silicon.
3. Diverse Customer Demand Outside the Big Clouds Limits Concentration Risk
Demand for high-end compute has moved far beyond a few giant cloud providers. The segment including national governments, regional cloud firms, and enterprises (which the company calls ACIE) represented roughly half of the data center business in the second quarter. This segment generated $40 billion in revenue during the quarter. That figure represents a 138% increase over the previous year. Sovereign AI efforts are particularly sticky because countries want to build local models for security and independence. These customers often prefer NVIDIA because its software ecosystem is already the global standard. This growing pool of buyers reduces the risk that a spending slowdown at one or two major tech firms could derail growth.
4. Strategic Acquisitions Extend Control Over the Model Development Ecosystem
The $12.9 billion acquisition of Hugging Face represents a decisive move to own the front door of AI development. This platform hosts over 3 million models and serves as the primary hub for millions of developers. By controlling the repository where open-source AI is built and shared, NVIDIA ensures that new models are optimized to run best on its own hardware. This strategy creates a powerful network effect where developers use the tools because they run best on the chips, and customers buy the chips because the software is already built for them. It's a software moat that protects the hardware business from rivals trying to compete on price alone.
5. Relentless Capital Returns Underpin Shareholder Value During High Growth
Management is utilizing its massive free cash flow to reward shareholders and cement its market lead. In the second quarter of fiscal 2027, the company returned $26 billion to investors. This included $20 billion in share buybacks and $6 billion in dividends. It still has approximately $99 billion remaining in its buyback authorization as of late July. This aggressive return of capital happens even as the company invests billions in its own supply chain and equity portfolio. Year to date, the firm has returned 60% of its free cash flow to stockholders. This financial discipline provides a solid floor for the stock while the company maintains its high-growth trajectory.
Potential Bear Cases
1. Rising Component Costs Threaten the Software Like Margin Profile
Skyrocketing prices for high-bandwidth memory are starting to erode the company's elite margin profile. Management warned that gross margins will likely bottom between 71% and 72% in the fourth quarter of fiscal 2027. This is a significant drop from the 75% margins reported in the most recent quarter. Memory scarcity is a side effect of the AI boom itself, as suppliers struggle to keep up with the demand for chips. These costs are expected to remain elevated well into next year. While the company is raising its own prices to compensate, there's a limit to how much of this inflation it can pass on to customers without impacting demand.
2. Persistent Supply Chain Bottlenecks Leave Massive Revenue Opportunities on the Table
The biggest threat to growth is a lack of parts. Management recently guided for 70% revenue growth in fiscal 2028. This figure is based on what the company can actually build and ship, not what customers want to buy. Internal forecasts suggest that demand is actually double what the current supply chain can fulfill. This means NVIDIA is leaving roughly 30 percentage points of potential growth on the table because it can't secure enough wafers and components. This bottleneck is expected to last through the end of fiscal 2028. Any further disruptions in the global supply chain could cause the company to miss these already constrained targets.
3. Structural Loss of the China Market Creates a Permanent Growth Ceiling
Geopolitical tensions have effectively erased the world's second-largest economy from the company's future growth plans. In the second quarter, data center revenue from China dropped to less than 1% of the total. Management has officially removed all China data center compute revenue from its forward outlook. This creates a permanent gap in the addressable market that other regions must work twice as hard to fill. Meanwhile, Chinese rivals are racing to build their own processors to replace the American technology they can no longer buy. This creates a long-term risk where the company loses its dominant position in a massive market and faces new global competition from state-backed alternatives.
The Moneyball Verdict: 99 out of 100
Our proprietary Moneyball scoring system assigns NVIDIA an overall Superscore of 99 out of 100, placing it in the Exceptional category. The Superscore is an AI-powered score that evaluates a company's overall strength by combining financial performance, product market position, technological capabilities, leadership quality, and relative valuation. It represents the unification of all our scores into a single score for public companies, with five rating bands: Exceptional (90-100), Strong (75-89), Above Average (60-74), Average (40-59), and Cautious (0-39). This score places the company in the Top ~1% of all companies we score, or ahead of roughly 99 out of every 100 companies in our database.
NVIDIA remains the dominant force because its transition to gigawatt scale AI factories creates a structural advantage that outpaces temporary margin and supply constraints. To doubt it, you'd have to believe that component costs or regional market losses will break a business model capturing $40 billion per gigawatt. This score suggests that the company's vertical integration is currently far too strong for those headwinds to matter.