Nvidia Hits $5.5 Trillion, Eclipsing Japan’s GDP in AI Infrastructure Concentration

Nvidia’s market capitalisation reached $5.5 trillion on 13 May 2026, making it the first publicly traded company to achieve the milestone and cementing the chipmaker’s dominance over global AI infrastructure at a scale that now exceeds Japan’s $4.4 trillion GDP and Germany’s $5.45 trillion economy.

The valuation, which hit an intraday peak of $5.72 trillion the following day, represents 23% of Japan’s economic output despite Nvidia employing roughly 30,000 people. The company now dwarfs the combined market value of all German and UK equities, according to NewsCase, placing it economically ahead of almost every country except the United States ($32.4 trillion) and China ($20.8 trillion).

The milestone exposes unprecedented capital concentration in AI compute layers, with Nvidia capturing the bulk of hyperscaler spending that Goldman Sachs projects will reach $7.6 trillion cumulatively between 2026 and 2031 across compute, data centres, and power infrastructure. The firm’s analysis assumes Nvidia will capture roughly 75% of compute spending, a dependency ratio that creates systemic risk as AI Infrastructure becomes strategic national infrastructure during US-China technology decoupling.

Revenue Trajectory Driven by Data-Centre Dominance

Nvidia’s data-centre segment generated $193.7 billion in fiscal 2026 revenue (ended 25 January 2026), representing 89.6% of total sales and a 68% year-over-year increase, per TheEnergyMag. Wall Street expects the segment to contribute around $73 billion in the upcoming Q1 FY2027 alone, with total quarterly revenue projected at $78.8 billion and adjusted earnings per share of $1.77.

The forward earnings multiple of 27.91x, reported by GuruFocus as of 14 May, sits well below Nvidia’s 10-year average P/E of 61.7x but masks the structural concentration risk embedded in the company’s revenue base. Management guidance projects $1 trillion in cumulative revenue from the Blackwell architecture and Vera Rubin platform by fiscal 2027, supported by hyperscaler commitments exceeding $500 billion through 2027.

Bank of America raised its total addressable market estimate for AI data centres to $1.7 trillion by 2030, up from a prior $1.4 trillion forecast, citing what Nvidia CEO Jensen Huang described as an “agentic AI inflection point” where computing demand is growing exponentially. 247 Wall Street noted the bank’s analysts expect hyperscaler capital expenditures to reach $725 billion in 2026 alone, driven by Meta, Amazon, Google, and Microsoft infrastructure buildouts.

China Export Controls Eliminate $20 Billion Market

The valuation milestone occurs against complete exclusion from China, where US export controls have reduced Nvidia’s market share to zero percent despite H200 chip export approvals that have yielded no actual sales. CEO Huang acknowledged the market exit during recent commentary covered by Tom’s Hardware, marking a collapse from the 66% share Bernstein estimated for 2024 to a projected 8% in coming years.

The geopolitical overhang carries material revenue implications. In Q2 and Q3 of fiscal 2025, China accounted for 13.11% and 16.92% of Nvidia’s revenue respectively, translating to $17.11 billion and $10.31 billion in sales. The elimination of direct access to the world’s second-largest economy occurs as Chinese AI model usage surged to 32% of worldwide token consumption in March 2026, up from just 5% a year earlier, according to Morgan Stanley data cited by TheStreet.

Grey-market channels remain active, with Nvidia B300 AI servers selling for approximately $1 million in China—nearly double US pricing—but the volumes cannot replace direct commercial relationships. The Institute for Progress estimated that allowing H200 exports would shrink the US AI compute advantage from 11-to-one to six-to-one by 2026, a strategic calculation that has frozen official sales despite regulatory approval.

Market Concentration and Circular Financing Patterns

The Magnificent Seven stocks—Nvidia, Apple, Microsoft, Alphabet, Amazon, Meta, and Tesla—now represent 35% to 40% of S&P 500 market capitalisation, up from historical averages that typically hovered in the low twenties. CNBC flagged the concentration as a structural vulnerability in late 2025, with Nvidia accounting for the largest single-stock weight increase during the AI infrastructure buildout.

The concentration coincides with circular financing dynamics, where hyperscalers purchase Nvidia chips to build capacity they may ultimately consume internally rather than lease to third parties. Jefferies analysts noted that “cap-ex continues to soar as demand outpaces supply and pricing increases,” but the sustainability of this cycle depends on AI application revenue eventually justifying the infrastructure investment. If utilisation rates disappoint or model training efficiency improves faster than expected, the feedback loop could reverse.

Valuation Sustainability and Competitive Threats

While Nvidia’s forward multiple sits below its 10-year average, the absolute market capitalisation creates mathematical constraints on future returns. A Motley Fool analysis comparing Nvidia to Alphabet in the race to $10 trillion noted that Nvidia “looks more likely to achieve a $10 trillion valuation first,” but Alphabet “has a better chance of maintaining that kind of market cap over the long run” due to business diversification and more sustainable valuation ratios.

Competitive pressure is mounting as hyperscalers develop proprietary AI chips. Amazon’s Trainium and Inferentia, Google’s TPU architecture, and Microsoft’s Maia chips all aim to reduce dependency on Nvidia’s ecosystem, though none have achieved performance parity with the latest Blackwell or Hopper generations. The threat lies not in immediate displacement but in gradual margin compression as customers gain negotiating leverage and workload optionality.

The geopolitical dimension adds non-commercial risk. Ainvest characterised China market access as “not a simple commercial deal but a geopolitical lever, and Nvidia is not currently in a position to pull it.” Export policy effectively weaponises Nvidia’s market position, creating both strategic value and regulatory overhang that traditional valuation models struggle to quantify.

What to Watch

Nvidia’s 20 May earnings call will test whether Q1 FY2027 guidance of $78.8 billion revenue holds amid Hyperscaler Capex scrutiny and utilisation rate questions. Watch for commentary on Blackwell production yields, customer concentration metrics, and any shift in China policy that could reopen the eliminated market segment.

Longer term, the sustainability of 75% compute share assumptions in Goldman Sachs’ $7.6 trillion infrastructure buildout depends on competitive positioning versus custom silicon and pricing power retention as hyperscalers negotiate volume discounts. The $1.7 trillion TAM forecast for 2030 AI data centres assumes continued exponential growth in model training and inference workloads—a trajectory vulnerable to efficiency breakthroughs or demand saturation.

Market Concentration remains the structural wildcard. If Magnificent Seven weighting approaches 50% of the S&P 500, index mechanics could force passive rebalancing or regulatory scrutiny. At $5.5 trillion, Nvidia has reached a scale where its valuation is no longer just a reflection of AI infrastructure demand but a determinant of broader market structure and geopolitical leverage in technology competition.