Nvidia's $3.8 Trillion Milestone: AI Dominance or Overvalued Bubble?

The tech world has rarely seen a valuation race as dramatic as the one NVIDIA is currently running. On June 2, 2025, the company's market cap surged to $3.35 trillion, fueled by a 78% year-over-year revenue jump to $39.3 billion in Q1 2025. Analysts at Loop Capital now project this figure could hit $6 trillion by 2028. Yet, as NVIDIA's stock price climbs—a —a critical question looms: Is this a testament to AI's transformative power, or the peak of an overvalued bubble?

The Rise of AI's Kingmaker

NVIDIA's ascent is undeniable. Its data center revenue, driven by AI infrastructure sales, hit $35.6 billion in Q4 2024, a 93% year-over-year surge. The company's Blackwell GPUs, critical for training large language models, have become the backbone of hyperscaler AI projects. Microsoft, Amazon, and Google all rely on NVIDIA's CUDA ecosystem, now boasting 4 million developers—a moat no competitor has yet breached.

But the real magic lies in margins. NVIDIA's net income hit $72.88 billion in fiscal 2025, a 145% jump from 2024. This has propelled its P/E ratio to 23x forward EV/EBITDA, far above AMD's 11.5x and Intel's 15x. The question isn't whether AI is real—it's whether NVIDIA can sustain this dominance as rivals close in.

The China Dilemma: A Thorn in the Side of AI's Growth

NVIDIA's growth faces a geopolitical speed bump. U.S. export restrictions on its H20 GPUs to China have already cost the company $2.5 billion in lost revenue in Q2 2025, with projections of an $8 billion hit in Q2 2026. China, once a critical market for AI chip sales, is now a no-go zone—a problem NVIDIA shares with AMD, which lost $1.5 billion in 2025 revenue due to similar restrictions.

While NVIDIA has launched compliant Blackwell variants, the long-term impact remains unclear. China's AI ambitions won't fade; they'll pivot to domestic suppliers like Baidu's Kunlun or AMD's MI400X chips. The result? A fragmented market where geopolitical risks could throttle NVIDIA's growth.

The Stealth Threat: AMD's Cost Advantage and Ecosystem Ambitions

AMD is no longer playing catch-up. Its EPYC Genoa CPUs now hold 39.4% of the data center market, up from 25% in 2023, and its MI300X GPUs undercut NVIDIA's A100 in cost while matching performance. Microsoft's choice of AMD's MI300X over NVIDIA's A100 for GPT-4 inference—a $4 billion project—signifies a turning point.

AMD's secret weapon? Full-stack integration. Acquisitions of Silo AI (enterprise software) and ZT Systems (rack-scale optimization) give it a 20–30% cost advantage in deploying AI clusters. By 2026, its MI400X GPU aims to match Blackwell's performance at half the price. Meanwhile, NVIDIA's software lead is eroding: AMD's ROCm 5.0 now supports 90% of CUDA workloads, weakening NVIDIA's ecosystem lock-in.

Valuation: A Tightrope Between Innovation and Overhang

NVIDIA's $3.35 trillion valuation hinges on two assumptions: that AI demand will grow exponentially, and that no competitor can replicate its ecosystem. But reality may be murkier.

• AI's Ceiling: While AI spending is projected to hit $150 billion by 2030, the current hype could fizzle if generative AI fails to deliver on its promise. NVIDIA's 23x P/E ratio assumes perpetual growth—a tough ask in tech.
• Competitive Pressure: AMD's valuation is half NVIDIA's despite its rising momentum. If AMD captures 20% of NVIDIA's data center revenue by 2026—a plausible scenario—NVIDIA's stock could correct sharply.
• Regulatory Risks: Antitrust scrutiny looms. The EU's Digital Markets Act could force NVIDIA to open its software stack to rivals, diluting its moat.

Conclusion: Position for Volatility, Not Certainty

Investors must parse reality from hype. NVIDIA's AI leadership is real, but its valuation demands perfection. The $3.35 trillion price tag assumes no major missteps in China, no breakthroughs from AMD, and endless AI demand.

For now, hold NVIDIA if you believe AI is a decade-long revolution and its ecosystem remains unassailable. But position for volatility: allocate no more than 5% of a portfolio to NVDA, and pair it with AMD (which trades at a 50% discount to NVIDIA's growth).

The $6 trillion bet by 2028? It's a gamble. AI's future is bright, but so are AMD's and Intel's ambitions—and the geopolitical clouds on the horizon.

Actionable Takeaway:
- Buy NVIDIA: If you're bullish on AI's long-term dominance and willing to tolerate geopolitical risks.
- Sell/Short: If you believe AMD's cost advantage and ecosystem integration will erode NVIDIA's margins.
- Wait: For clarity on China's AI strategy and the durability of large language model adoption.

The race for AI supremacy isn't over—but the finish line is crowded with rivals, and NVIDIA's lead is narrowing.