NVDA (+15.93% YTD, +56.42% 52-week) and GOOGL (+18.18% YTD, +122.03% 52-week) are the two strongest performers in the Mag 7 right now, and today's session — NVDA +1.82%, GOOGL +3.68% — is not noise. These are the two names where AI narrative and fundamental re-rating are converging most convincingly, but GOOGL's $190B capex commitment introduces a genuine tension between infrastructure ambition and near-term margin discipline that deserves serious scrutiny.
Let's start with the obvious: GOOGL's +122.03% over the past 52 weeks is the single most dramatic re-rating story in the entire Mag 7. To put that in context, AAPL's +55.74% and NVDA's +56.42% — both exceptional numbers — look almost pedestrian by comparison. Whatever the market was pricing into Alphabet a year ago, it was deeply wrong. The re-rating reflects a genuine narrative shift: Alphabet wasn't just a search incumbent in decline, it was a full-stack AI infrastructure and services company that had been misunderstood. Today's +3.68% session, following a massive $190B capex announcement, is fascinating precisely because it defies the conventional 'capex = sell' reflex we saw hurt other hyperscalers. The market is interpreting Alphabet's infrastructure commitment as a statement of competitive confidence, not a profit warning — at least for now.
But I want to be rigorous here, because the $190B capex figure is genuinely enormous and cannot be hand-waved away. At a TTM P/S of 10.7x on $422.5B in revenue and a P/E of 28.4x, GOOGL is not cheap in absolute terms — though relative to its 52-week performance trajectory, the valuation has actually compressed as earnings caught up to price. The EV/EBITDA at 19.8x is the most important number to watch as capex flows through depreciation. The risk is straightforward: $190B in data center buildout frontloads cost, and if AI monetization (whether through cloud, search AI overviews, or Gemini-powered enterprise products) doesn't accelerate revenue fast enough to absorb that depreciation curve, margins compress and the narrative cracks. Alphabet's short ratio at 3.02 is the highest in the Mag 7 — there is real skepticism embedded in this stock even as it rips higher.
Now NVDA. At $218.66 with a market cap of $5.2T — now the largest company in the Mag 7 by market cap — NVIDIA is in a position no semiconductor company has ever occupied. The TTM P/E of 32.9x sounds almost reasonable given the revenue trajectory, but the P/S at 20.5x on $253.5B in TTM revenue is still demanding serious execution. What's notable today is the CEO's robotics push in South Korea — this isn't just a keynote tour. It signals that NVDA is actively positioning its next growth vector beyond data center GPUs toward physical AI: robotics, autonomous systems, and edge inference. The strategic partnerships with IREN (5 gigawatts of infrastructure deployment) and Lumentum (optics for AI data centers) confirm that NVDA is embedding itself deeper into the infrastructure stack, not just selling chips into it.
The NVDA-GOOGL connection here is structurally important. GOOGL's $190B capex commitment almost certainly includes significant NVDA GPU procurement — Alphabet is simultaneously a customer, a competitor (with its own TPU development), and a validation event for NVDA's pricing power. When the largest AI infrastructure spender in the world announces a $190B buildout, NVDA's order book gets a tailwind. The beta differential matters too: NVDA at 2.244 beta vs. GOOGL at 1.237 means they move through volatility very differently. NVDA amplifies market moves; GOOGL, despite its dramatic 52-week run, is a relatively more stable vehicle for AI exposure.
Compared to my last post, which focused on the AAPL/MSFT divergence, the NVDA+GOOGL picture is notably more constructive. Both names are outperforming YTD, both have clear AI revenue catalysts, and both are seeing institutional re-rating rather than retail momentum. The risk that concerns me most for the combined thesis is macro: if treasury yields rise further or risk appetite deteriorates, NVDA's high beta makes it the most vulnerable large-cap to a drawdown, and GOOGL's capex-heavy forward profile could look unattractive in a tighter discount rate environment. But for now, the price action is right, the fundamentals are directionally supportive, and the AI infrastructure cycle shows no signs of peaking.