Just caught something worth paying attention to. Nvidia's Jensen Huang basically just told investors that the AI infrastructure spending opportunity is way bigger than most people realize, and the company's next-gen Vera Rubin platform is positioned to dominate it.

Here's the thing - during their latest earnings call in late February, Huang made this point that stuck with me. The world historically spent around $400 billion annually on classical computing infrastructure. But AI workloads? They need roughly a thousand times more capacity. Let that sink in for a moment.

The Vera Rubin platform is no joke either. We're talking about GPU architecture that can train AI models using 75% fewer chips compared to the current Blackwell generation, while cutting inference token costs by 90%. That's the kind of efficiency gain that changes economics across the entire industry. Nvidia's already got samples shipping, with mass production ramping in the second half of this year.

Looking at the numbers, Nvidia just posted $215.9 billion in revenue for fiscal 2026 - that's a 65% jump year-over-year. Data center sales alone hit $193.7 billion. And Jensen Huang's not wrong about where this is headed. The company's guiding for $78 billion in Q1 FY2027 revenue, which would represent 77% growth. Most of that's coming from data center.

What's interesting from a valuation perspective - the stock is trading at a P/E of 36.1 right now. That's actually 41% below its 10-year average of 61.6. Forward P/E sits at just 21.5 based on Wall Street's fiscal 2027 earnings estimate of $8.23 per share. For comparison, the S&P 500 trades at a trailing P/E of 24.7.

So you've got this situation where Jensen Huang is basically laying out a trillion-dollar opportunity thesis, the company's demonstrating execution with this next-generation platform, and the stock's trading at a significant discount to its historical valuation. Whether that gap closes is probably the real question here.