Nvidia, the world's largest company by market capitalisation, is scheduled to report earnings on 26 August. The results are expected to provide insight into whether the AI investment boom is broadening or becoming dependent on a small number of firms and products.

In May, Nvidia reported $81.6 billion in revenue, with $75.2 billion originating from its data centre business. The company provides semiconductor chips used by frontier AI labs and major tech companies to build computing capacity.

Data centre revenue and product demand

S&P Global Market Intelligence has set a consensus estimate for data centre revenue at $85.7 billion, with a range between $83.5 billion and $91.5 billion. Total consensus revenue is estimated at $92.2 billion. Analysts are monitoring how usage patterns evolve, specifically the migration of capacity consumption toward inference workloads rather than training.

The adoption of Groq 3 LPX, Nvidia's implementation of technology optimised for inference, will be a factor for future guidance. As agentic AI use cases increase, demand may shift from specialised GPUs to include traditional CPUs, where Nvidia faces competition from Intel, Arm, AMD, and chips created with Groq's IP.

The Rubin transition

Nvidia's next-generation compute platform, Rubin, is expected to begin shipping in the fall. Morgan Stanley estimates Rubin will contribute nearly $9 billion in the quarter ending in October. The company claims the new chip offers up to 10 times lower inference cost per token compared to current technology.

The transition from the Blackwell platform to Rubin presents potential risks. Nvidia has previously warned that product transitions can "result in revenue volatility" because "customers may postpone purchasing new architectures or may adopt new technologies more gradually than anticipated."

Customer concentration and financing

Nvidia's quarterly filing for the period ending in April 2026 shows revenue is concentrated among a limited number of direct and indirect customers. Three customers accounted for:

• 21% of revenue
• 17% of revenue
• 16% of revenue

These three entities also accounted for 64% of total accounts receivable. Analysts are looking for evidence of demand spreading to enterprises, industrial users, sovereign projects, and private AI clouds.

The financing of AI infrastructure is moving through several stages:

• Balance sheet investment: Large players using free cash flow, such as Microsoft's $10 billion investment in OpenAI in 2023.
• Debt and combined capital: OpenAI's 2025 Stargate project, which uses equity, sovereign capital, and debt for a projected $500 billion investment. Oracle, a key infrastructure provider for the project, was reported by the Wall Street Journal to be straining under significant debt by mid-2026.
• Equity capital raises: Alphabet announced an $80 billion equity capital raise in June following a quarter of negative cash flow driven by capital expenditures.

Nvidia has entered an agreement with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to create a financing platform for more than $500 billion dedicated to AI infrastructure. CEO Jensen Huang described the strategy by stating:

"In AI, compute is revenue. NVIDIA compute is uniquely suited for this role."

Brian Mulberry, chief market strategist at Zacks Investment Management, noted to Yahoo Finance that

"Nvidia is kind of acting as the central bank of AI."

Wharton professor João Gomes, writing for Fortune, warned that

"Financial vulnerabilities can remain hidden until they become crises."