NVIDIA just released a set of earnings that, on paper, looked like a “beat across the board.” Revenue and EPS topped analyst forecasts, gross margin held strong, and the buyback program was boosted significantly. But despite all that, the stock dropped nearly 4% after hours, slipping from $181 to $175, and the market paid close attention.
Source: Google Finance
So why would the hottest company in the AI era stumble right after posting such strong numbers? The answer lies in more than just slowing growth in its core data-center business. Investors are asking the bigger question: when expectations are already maxed out and valuations stretched, can NVIDIA still keep delivering the “future” everyone has priced in?
The Monopoly on AI Compute Power
Right now, NVIDIA is the undisputed king of AI chips, with roughly 97% market share in training GPUs. AI progress ultimately depends on three things: data, compute, and algorithms—and compute is the decisive one.
NVIDIA doesn’t just sell the fastest hardware; it’s built the CUDA ecosystem that locks developers into a full stack of software + hardware, creating a moat that’s hard to cross. Rivals like AMD and Intel are trying to catch up, but in the high-end segment, they still can’t mount a real challenge.
Breaking Down the Q2 Report
1. Strong Results, but Growth Slowing
-
Revenue came in at $46.74B, up 56% YoY, beating Wall Street’s $46.23B forecast and even topping the company’s own guidance high end ($45.9B).
-
EPS hit $1.05, up 54% YoY and above the $1.01 expected.
-
Ex-H20 adjustments, gross margin was a stable 72.3%.
The catch? Growth is clearly decelerating. Last quarter revenue grew 69% YoY; this quarter it’s down to 56%—the slowest pace in more than a year. That has investors asking whether the AI spending peak is already behind us.
2. Core Businesses Diverging
-
Data Center: Revenue of $41.1B, up 56% YoY but down 1% sequentially. H20 shipments fell off, cutting ~$4B of sales. Blackwell chips grew 17% sequentially, networking almost doubled YoY, but momentum is slowing overall.
-
Gaming & AI PCs: Revenue of $4.3B, up 49% YoY and hitting an all-time high, thanks to the AI-PC upgrade cycle and stronger Blackwell shipments.
-
Professional Visualization: $601M, up 32% YoY, slightly above expectations.
-
Auto & Robotics: $586M, up 69% YoY but short of forecasts, showing softer momentum.
Source: gurufocus
3. Management’s Cautious Outlook—China in Focus Guidance for Q3 is $54B in revenue—above consensus, but well below the most bullish ~$60B forecasts. Management is clearly managing expectations.
CEO Jensen Huang highlighted that China could represent a $50B opportunity this year, and confirmed export approval for the H20 chip. But Q2 included no shipments to China, and Q3 guidance doesn’t assume any either. That leaves U.S.–China policy as the biggest external swing factor.
Huang also stressed that NVIDIA isn’t just a “chip company” but a full AI systems provider, with product moat and efficiency advantages. Still, he admitted custom ASICs could bring structural competition.
Meanwhile, the board approved an extra $60B share buyback, signaling long-term confidence—though the stock is already trading at record valuations.
The Risk of Expectations Running Too Hot
NVIDIA’s stock has surged to all-time highs, not only because of strong fundamentals and tech leadership, but because of extreme optimism about AI’s future. That “over-crowded consensus” is itself a risk.
Today’s valuation assumes almost a flawless future: demand always full, AI spending always rising, and tech transitions frictionless. That’s a tall order. Even a slight slowdown in growth can trigger sharp corrections when expectations are this stretched.
The other risk: NVIDIA has become the ultimate crowded trade. Both passive and active funds are heavily overweight, making the stock seem “unshakable.” But when positioning is this concentrated, prices don’t need bad news to fall—they just need expectations to reset.
For context, NVIDIA alone has contributed more than 25% of the S&P 500’s total gains in 2024. When a single company carries that much market weight, even tiny changes in fundamentals get magnified into big swings in sentiment.
It’s not that NVIDIA is doing worse—it’s that expectations have gone so far ahead, the stock naturally runs into gravity.
Invesight Viewpoint
The real question isn’t “Is NVIDIA a good company?”—it clearly is. The question is: with valuations sky-high and positioning overcrowded, is it still worth chasing here?
My take: NVIDIA remains one of the most strategically important tech companies in the world. But at these levels, buying into it is less about fundamentals and more about betting against expectations themselves.
Comments