Nvidia Reassures on Roadmap and Broadening AI Demand, Citi Says

Nvidia Corp. has affirmed that its AI product roadmap remains on schedule and that demand for its accelerators is broadening beyond the initial wave of hyperscale cloud customers, according to a recent analyst note from Citi following discussions with the chipmaker's investor relations team.

Roadmap and Demand Confirmed

Nvidia told Citi that its future product pipeline, including the Rubin Ultra platform, remains fully intact with no changes to its NVLink domain architecture. This addresses investor questions about the company's ability to maintain its rapid innovation cadence amid intensifying competition.

The company also highlighted a significant expansion in its customer base. While hyperscalers were the primary drivers of initial AI infrastructure spending, Nvidia said demand has gained momentum over the past two years from other key segments.

• AI laboratories
• Sovereign AI customers (government-backed initiatives)
• Enterprise on-premise deployments

Nvidia expects these segments to represent a larger share of the market as AI adoption accelerates, though it declined to comment specifically on spending plans from Meta Platforms, according to the note.

Financials and Efficiency

Ad

Management reiterated its long-term gross margin target in the mid-70% range and confirmed its plan to return 50% of cash flow to shareholders this year, noting that buybacks could increase over time. The company's recent $25 billion debt offering was positioned as a move to increase financial flexibility.

Nvidia also clarified recent comments from CEO Jensen Huang regarding future energy efficiency. The projection that $100 billion of compute could eventually require just one gigawatt of power reflects expected technological gains, not current infrastructure costs. The company estimates its current systems generate roughly $30 billion to $40 billion in compute revenue per gigawatt, with newer generations like Blackwell offering significant power efficiency improvements over the Hopper architecture.

Analyst Perspective

Following the discussion, Citi reiterated its rating for Nvidia as its top buy-rated mega-cap data center semiconductor stock. The bank cited Nvidia's strong access to constrained high-bandwidth memory (HBM) supplies as a key competitive advantage.

Nvidia also stated that its in-house AI models, such as Nemotron, are designed to support enterprise and sovereign adoption rather than to compete directly with leading frontier-model developers. The company said it has not officially announced a collaboration with AI chip startup d-Matrix but remains open to integrating third-party technology into its platform.