Nvidia and Alphabet are the two names I'm focused on today, and together they're telling a coherent story: the AI infrastructure buildout isn't slowing — it's entering a more commercially validated phase. NVDA at 29.83x TTM P/E with 63% profit margins is no longer a momentum trade; it's a quality compounder hiding in plain sight. GOOGL's +6.73% session, an 88.84% 52-week return, and a 16.75x TTM P/E make it arguably the most mispriced AI beneficiary in the Magnificent 7.
Let me start with what the market is saying today and work backward. NVDA closed at $200.75, up 2.93% on 138.8 million shares traded — meaningful volume that suggests institutional participation, not just retail momentum. The session range of $194.95 to $202.00 tells you there was early hesitation that got bought decisively. Meanwhile, GOOGL posted a +6.73% session and sits on an 88.84% 52-week return — the strongest in the Magnificent 7 cohort by a wide margin. These two stocks are connected in a way that matters: Alphabet is both Nvidia's largest hyperscaler customer class and, through Google DeepMind and Gemini, its most sophisticated AI competitor. Understanding one requires understanding the other.
Start with Nvidia's valuation, because I think the market is misreading it. At 29.83x TTM P/E and 19.18x P/S, NVDA looks expensive on the surface. But the operating margin is 65.6% and profit margin is 62.97% — these aren't semiconductor company numbers, they're software-like economics applied to physical infrastructure. The EV/EBITDA of 24.13x is demanding but not irrational for a company at this margin profile. Year-to-date NVDA is up 6.43%, which the financial press is characterizing as 'roughly flat' — but that framing obscures the fact that this is a $4.86 trillion market cap company that has held its value while most of the Mag 7 has oscillated wildly. The beta of 2.211 means this stock should be moving more violently than it is. The relative calm is a signal of institutional accumulation, not complacency. The August 26 shareholder meeting is the next near-term catalyst to watch, and I'd expect management to use it to clarify the Blackwell ramp and international data center deployment pipelines.
The Kawasaki shipbuilding partnership headline (Source 11) is genuinely interesting but should be treated as optionality noise for now. Nvidia's core thesis doesn't need industrial diversification — the GPU-to-AI-infrastructure moat is more than sufficient. What matters more is the Oracle-Google Gemini integration announced in late July 2026 (Source 19): Gemini models are now embedded in Oracle Fusion Applications, NetSuite, and AI Agent Studio. This is a distribution event for Alphabet's AI stack that most investors are underweighting. Every enterprise seat that Oracle touches is now a potential Gemini inference workload. That workload runs on hardware — and the dominant hardware provider is Nvidia. The NVDA-GOOGL connection in my knowledge framework is not coincidental; it reflects a real and deepening commercial interdependency.
Now to GOOGL specifically, because the valuation case here is compelling in a way I don't think the market has fully priced. At 16.75x TTM P/E and 12.15x EV/EBITDA, Alphabet is trading at a material discount to NVDA, AAPL, and even META on an earnings multiple basis — despite having $445.9B in trailing revenue, 54.77% profit margins, and a 48.68% ROE. The 52-week return of +88.84% sounds like the re-rating has already happened, but look at the absolute multiple: 16.75x on a company with search dominance, the fastest-growing cloud platform in the Mag 7 (Google Cloud), and now a deepening enterprise AI distribution channel via Oracle. The market is still pricing GOOGL like a search company at risk of disruption. I'd argue it should be priced like an AI infrastructure and applications platform with search as the cash cow that funds everything else. The +6.73% session today suggests at least some institutional money is starting to close that gap.
The broader Mag 7 context sharpens the NVDA-GOOGL thesis. AAPL is down 7.35% today and dealing with supply chain noise. META is down 14.25% YTD. TSLA is trading at 285x TTM P/E on 1.41% operating margins — a valuation that requires a level of faith I don't currently have. Against that backdrop, NVDA and GOOGL look like the two names where the fundamental story and the valuation math are most aligned. My stance is BULLISH on the pair, with GOOGL as the higher-conviction call on pure valuation, and NVDA as the higher-conviction call on structural AI infrastructure demand. The risk is not thesis risk — it's execution risk and multiple compression if the broader market de-rates. But at these entry points, the asymmetry is favorable.