NVDA at $222.82 is modestly off today (-0.69%) but the real story is Jensen Huang commanding the AI infrastructure conversation from Taipei — optical connectivity is the next bottleneck narrative, and NVDA is positioned squarely at the center of it. GOOGL at $361.85 (-3.86% today) is taking a harder hit, but an $80B buyback and a 52-week return of +118.50% tell you the market isn't abandoning the thesis — it's digesting it. I'm MIXED: structurally bullish on NVDA's infrastructure dominance, constructively cautious on GOOGL's near-term price action despite compelling capital return signals.
Let's start with NVDA, because Jensen Huang doesn't take a stage without a purpose. His remarks at the Taipei tech event this week — specifically flagging optical connectivity as the critical bottleneck for next-generation AI infrastructure — were not casual observations. This is deliberate narrative construction. By identifying the next chokepoint in the AI buildout, Huang simultaneously validates the ongoing capex supercycle that drives NVDA's datacenter revenue AND positions the company as the oracle of what comes next. The Marvell endorsement ('next trillion-dollar company') was a secondary signal, but an important one: NVDA's CEO is now actively shaping the competitive landscape by elevating ecosystem partners, which reinforces NVDA's role as the gravitational center of AI infrastructure. With NVDA trading at $222.82, a P/E of 34.1x, and a 52-week return of +57.82%, the valuation is demanding but not irrational given the revenue trajectory — TTM revenue of $253.5B with an operating margin north of 65% is a fundamentals profile that very few companies at any market cap have ever produced. The beta of 2.244 is a real risk factor, particularly with an inflation print due in roughly a week, but the structural demand case remains intact.
On GOOGL, today's -3.86% move deserves careful reading. The $80B buyback announcement — characterized by Goldman Sachs as 'unprecedented' — is one of the most aggressive capital return signals in Big Tech history, and it's happening at a stock that has already returned +118.50% over the past 52 weeks. That combination of price appreciation AND a massive buyback authorization tells you management believes the stock is still undervalued at current levels, even after a historic run. The BofA caution flag on large-cap AI names including GOOGL is worth noting — valuation re-rating risk is real when expectations are already elevated — but GOOGL's TTM revenue of $422.5B at a P/E of 27.6x and EV/EBITDA of 20.6x actually looks more digestible than several Mag 7 peers. The space-based AI data center angle (Google exploring orbital infrastructure alongside SpaceX) is speculative but signals that Alphabet is thinking about compute capacity constraints in a genuinely differentiated way.
The NVDA-GOOGL connection matters more than most investors appreciate. NVDA's GPU infrastructure is a primary input cost for Alphabet's AI buildout — every dollar Google spends on Gemini training and inference is a dollar that flows through the NVDA supply chain. When Jensen Huang speaks about optical connectivity bottlenecks at a Taipei tech event, Google's infrastructure team is in the audience (figuratively if not literally). This symbiosis is a key reason why both stocks have dramatically outperformed over the past year, and it's a dynamic that doesn't reverse unless either demand signals from hyperscalers cool materially OR a credible alternative accelerator architecture gains commercial traction. Neither condition appears imminent.
From a positioning standpoint, NVDA remains my highest-conviction Mag 7 name for a 12-month horizon. The TTM revenue base is growing, the margin structure is exceptional, and the CEO is actively managing the long-term narrative with precision. For GOOGL, the near-term setup is choppier — a -3.86% down day on top of an already elevated 52-week return invites the question of whether institutional players are trimming into strength ahead of the inflation print next week. But the $80B buyback creates a structural buyer of last resort, and at 27.6x earnings with this revenue scale and AI optionality, I wouldn't be a seller. MIXED stance overall, with NVDA as the clear leader of the pair.