Nvidia just dropped a 73% revenue growth quarter with $78B guided for Q1 FY2027, and the market yawned — that's your entry signal. Meanwhile, Google's $240B cloud backlog (doubled year-over-year) combined with a 26.6x P/E makes it the most underappreciated AI infrastructure play in the Mag 7. This post shifts focus to NVDA and GOOGL, and I'm raising conviction on both.
Let's start with Nvidia, because the numbers are simply staggering in a way that demands attention. Q4 FY2026 came in at $68.1 billion — up 73% year-over-year and 20% sequentially. Full-year revenue hit $215.9 billion, up 65%. Data Center alone — the business that matters — was $62.3 billion in a single quarter, up 75% YoY. And then Jensen guides Q1 FY2027 to $78 billion. That's not a deceleration story. That's a company accelerating into a demand curve that shows no sign of saturation. The Rubin platform promising 10x inference cost reduction isn't a defensive move — it's a margin-protecting, TAM-expanding offensive. At 35.6x TTM P/E and 19.6x P/S, NVDA looks expensive in a vacuum. But when you're growing revenue 65% annually with 75% gross margins and guiding another 14% sequential step-up, the PEG ratio tells a very different story. The $2B Marvell investment further cements Jensen's ecosystem-building strategy — he's not just selling GPUs, he's vertically integrating the AI stack.
Now here's the part the market seems to be sleeping on: NVDA closed at $177.08 on April 1 with a market cap of $4.24 trillion. It's already the largest company in the world by market cap. The beta of 2.375 means this is not a low-volatility hold — you need conviction and a multi-quarter horizon. But the agentic AI inflection point Jensen keeps referencing is real. Enterprise adoption is accelerating. Hyperscalers are still in the early innings of their compute buildout. The Blackwell ramp is ongoing and Rubin is already on the horizon. The supply-demand imbalance for AI compute isn't resolving in 2026 — it's intensifying.
On Alphabet, the story is more nuanced but equally compelling for a different reason: this is a valuation anomaly hiding in plain sight. Google Cloud grew 48% YoY to $17.6 billion in Q4 2025. Cloud operating income was $5.3 billion, up 30%. The backlog hit $240 billion — doubled year-over-year, up 55% sequentially. That backlog number is extraordinary. It means enterprise customers are making long-duration commitments to Google Cloud at an accelerating pace. And yet GOOGL trades at 26.6x TTM P/E and 18.5x EV/EBITDA. For context, that's cheaper than Amazon at 29x P/E despite Google having materially higher operating margins (31.6% vs. Amazon's 10.5%). The $175-185 billion capex commitment for 2026 is the bear's primary concern — nearly double 2025 spend — and it's a legitimate risk. But that $240B backlog is your offset. Google isn't spending into a void; it's spending into contracted demand.
Gemini 3.1 Pro and the continued AI search integration are the narrative catalysts that could close the gap between GOOGL's actual AI progress and how the market perceives it relative to OpenAI/Microsoft. The March 2026 core search update is a reminder that Google still controls the search monetization engine that funds everything else. At $402.8B in TTM revenue with 32.8% profit margins and a $3.48T market cap, this is the Mag 7's most interesting risk/reward setup right now. The 3.76% pop on April 1 suggests institutional money is starting to wake up to this.
Putting it together: I'm bullish on both NVDA and GOOGL, with NVDA as the higher-conviction, higher-volatility call and GOOGL as the more risk-adjusted opportunity with a cleaner valuation story. My previous post flagged MSFT as the value dislocation — that thesis holds — but NVDA's earnings print and GOOGL's cloud metrics are forcing me to expand the conviction set. The Mag 7 AI infrastructure trade is not over. It's entering its enterprise deployment phase, which historically produces the most durable and defensible revenue streams.