NVIDIA CEO Jensen Huang just told us what the market already knew but few wanted to price in: we need trillions more dollars for AI infrastructure buildout. Not billions. Trillions. His timing is pointed, heading to China as US-China tech competition intensifies and global data center capacity maxes out.
Huang laid out the path: “I believe that there will be no digestion until we modernize a trillion dollars with the data centers… And let’s say by 2030, the world’s data centers for computing is a couple of trillion dollars.” NVIDIA is capturing 80-85% gross margins on the infrastructure powering this transformation.
Here are five stocks positioned to profit from this multi-trillion dollar wave, ranked by opportunity and risk.
5. Meta Platforms: The Vertical Integration Play
META is spending $18.8 billion per quarter building its own AI infrastructure rather than renting compute from cloud providers. The market is punishing it—down 8.5% year-to-date and flat over the past year despite strong earnings beats.
Meta reported $2.7 billion in net income for Q3 2025, artificially depressed by an $18.9 billion tax expense. Strip that out and you see healthy profitability, but the capex burn is real. Reality Labs continues bleeding cash while the company builds data centers.
The bull case depends on whether Meta can monetize this infrastructure faster than competitors buying compute as needed. At a forward PE of 21x (cheapest in this group), you’re betting on that monetization timeline accelerating. If you believe Meta’s AI investments will drive meaningful revenue growth by 2027, this is your entry point.
4. Amazon: AWS Must Deliver
AMZN has the highest quarterly capex at $35.1 billion, and investors are betting AWS can pull the stock out of its malaise. The company is flat year-to-date and up just 4.7% over the past year, dramatically underperforming other infrastructure plays.
AWS is the profit engine, but retail creates noise in earnings. Q3 2025 showed $21.2 billion in net income on $180.2 billion in revenue, with an 11.1% profit margin. The forward PE of 30x prices in meaningful growth that must come from AWS monetizing AI workloads.
Amazon’s advantage is existing customer relationships. If you’re already on AWS, adding AI compute is friction-free. The risk is customers increasingly want specialized AI chips rather than paying premium prices for NVIDIA-powered instances. Buy Amazon if you believe AWS maintains its enterprise stranglehold and successfully monetizes the AI transition.
3. Microsoft: The Azure Bet
MSFT spent $19.4 billion on capex in Q1 FY2026, almost entirely for Azure data center buildout. The stock is down 6% year-to-date despite 18.4% revenue growth, suggesting the market worries about returns on this massive infrastructure spend.
Microsoft’s advantage is the enterprise software stack. Azure isn’t just compute—it’s the platform running Office 365, Dynamics, and Copilot AI features. Q1 FY2026 showed $27.7 billion in net income on $77.7 billion revenue, with a 48.9% operating margin. That’s world-class profitability even while funding this buildout.
The forward PE of 29x is reasonable if Azure AI revenue accelerates. Microsoft has beaten earnings estimates for four consecutive quarters by an average of 5.3%. The risk is that Google’s TPUs or custom chips from Meta and Amazon erode Azure’s competitive position. Buy Microsoft if you believe the enterprise software moat makes Azure the default AI platform.
2. AMD: The Alternative Play
AMD just locked in a 10-year, $311 million data center lease with Riot Platforms, with potential to reach $1 billion with extensions. That’s validation that AMD’s MI455X accelerators are credible alternatives to NVIDIA’s chips, especially for customers wanting to avoid single-vendor lock-in.
The stock is up 96% over the past year and 8.3% year-to-date. Q3 2025 delivered $1.2 billion in net income on $9.2 billion revenue, with a 51.7% gross margin—solid but nowhere near NVIDIA’s 73.4%. The forward PE of 37x prices in significant growth, and analysts have a consensus target of $286 (38% upside).
AMD’s advantage is being the number-two player in a market where customers desperately want competition. Meta was AMD’s largest AI GPU customer in 2025. KeyBanc notes that AMD’s server CPUs are nearly sold out for 2026. The risk is that NVIDIA’s software ecosystem (CUDA) remains too entrenched and AMD captures share but not margins. Buy AMD if you believe data center customers will diversify chip vendors to avoid NVIDIA dependency.
1. NVIDIA: The Infrastructure King
NVDA is the obvious leader, and Huang’s China visit signals global expansion of the AI infrastructure buildout. The stock is up 33% over the past year despite being down 4.5% year-to-date, showing recent consolidation after a massive run.
Q3 FY2026 numbers are staggering: $31.9 billion in net income on $57 billion revenue, with a 73.4% gross margin and 63.2% operating margin. TTM revenue hit $187.1 billion with a 53% profit margin and 107% return on equity. The forward PE of 24x is the lowest in this group, suggesting the market expects continued explosive growth.
NVIDIA’s advantage is the complete stack: chips, networking, software, and increasingly, cooling solutions for next-gen data centers. The China visit matters because it signals NVIDIA is navigating geopolitical constraints while expanding globally. Jefferies just raised their price target to $275, and billionaire hedge fund managers (Asness, Englander, Cohen) bought shares in Q3 2025.
The risk is that growth moderates and the multiple compresses. But if Huang is right about trillions more needed, NVIDIA captures the largest share of that spending. This is the safest bet in the AI infrastructure wave and the most liquid way to play the thesis.
Top Gaining Stocks
Top Losing Stocks