Before buying a stock, ask yourself one question: does management deserve your capital?
KEY TAKEAWAYS
NVIDIA's entry into the AI PC market triggered a sharp market reaction, with NVIDIA, Arm, Dell, and HP rising while Intel, AMD, and Qualcomm declined. Investors viewed the announcement as a potential shift in the future competitive landscape of personal computing.
CUDA remains one of the most important competitive advantages in technology. The software ecosystem, developer adoption, and accumulated expertise create barriers that are significantly harder to replicate than hardware specifications.
The launch reflects years of strategic planning rather than a single product decision. Partnerships with Microsoft and MediaTek helped solve historical Windows-on-Arm challenges that previously limited adoption.
Successful market leadership requires both long-term strategic vision and short-term tactical execution. Identifying industry bottlenecks and eliminating them can be more valuable than introducing new technology alone.
Management quality remains a critical factor in long-term corporate success. Companies that consistently allocate capital well, anticipate market shifts, and execute effectively often create durable shareholder value.
MY HOT TAKES
The most valuable assets in technology are increasingly ecosystems rather than products. Software platforms, developer communities, and network effects can create competitive advantages that endure for decades.
Investors often underestimate the importance of management quality when evaluating companies. Leadership decisions made years earlier frequently determine which firms emerge as industry winners.
The market rewards evidence of execution more than innovation alone. New products matter, but demonstrating the ability to repeatedly deliver on strategic objectives matters even more.
Many failed technology initiatives share a common characteristic: they address symptoms rather than root causes. Sustainable success often comes from identifying and solving the underlying reasons previous efforts failed.
Long-term investing success frequently depends on identifying companies led by operators who can balance patience with decisive action. Strategic vision without execution rarely creates lasting value, while execution without vision often leads to stagnation.
You can quote me: “The market didn't reward a chip launch yesterday–it rewarded a decade of management execution.
It’s all about the base. On June 1, Jensen Huang walked into Computex and did something NVIDIA had never done before: he built a PC chip. And then he dropped the mic–in true rockstar flex mode. By the closing bell, the market had already handed down its verdict–and it wasn't close.
NVIDIA rose. Arm ripped higher. Dell and HP, the laptop makers who'd signed on to build the things, surged. And on the other side of the ledger, the incumbents… well, bled: Intel, AMD, and Qualcomm all fell as a single keynote rewrote the question every PC buyer will face this fall.
Here's the thing about that scoreboard. It looks like NVIDIA got lucky–right chip, right moment, right AI wave. My friends, it wasn't luck. It was the payoff of a decade of decisions: a software moat called CUDA, a three-year quiet collaboration with Microsoft, a CPU partnership that de-risked the one piece NVIDIA didn't have.
My longtime followers know that I always put “management” on the top of the list when describing a company. Without credible management that demonstrates skill in execution, strategy, and most importantly making tactical calls in a competitive, fast-moving market, a company–no matter how great its product–has a low probability of surviving, long-term.
So, let's talk about how Huang actually pulled this off, because the “how” is the whole story. Start with the moat. If you are a tech nerd, you would recognize that RTX Spark is a gorgeous piece of silicon. 🤓 It’s a Blackwell GPU fused to a 20-core Arm CPU with a frankly absurd 128GB of unified memory. 🐎 That’s tech spec, but the chip isn't the weapon. The weapon is CUDA, the software layer NVIDIA has been quietly nurturing for more than fifteen years. Anybody with enough cash and a slot at TSMC can fab a fast chip. Nobody can clone fifteen years of developers who already know how to build on your platform. That's not a product, that's a base. Huang started pouring that foundation back when "AI PC" wasn't even a phrase yet, and on June 1 he simply strolled out and stood on top of it like he owned the place. Which, increasingly, he does.
Then there's the patience, and patience is the part nobody in social media investing has much of a stomach for. NVIDIA didn't try to bootstrap a world-class CPU from a cold start. It leaned on MediaTek (a Taiwanese fabless semiconductor company on nobody’s radar) for the Arm cores and spent three years working hand-in-glove with Microsoft before anyone outside the room knew the project existed. That is strategy in its purest form: deciding years ahead which fight you're going to pick, then assembling the pieces so quietly that the incumbents never get a chance to prepare. By the time Intel realized it was in a war, the war was already being livestreamed to the planet.
But strategy alone doesn't win anything, and this is where the tactical genius hides–the part most people will skim right past. Windows-on-Arm is not a new idea. It has actually died before, repeatedly, and, frankly, humiliatingly. QUALCOMM planted that flag for years and never broke through. Why? Because the apps didn't run properly. The dream kept face-planting into the same two rocks. Huang's team looked at that graveyard and, instead of marching in with a shinier chip and a prayer, went and fixed the actual causes of death. Microsoft fixed its OS so x86 software runs cleanly, and NVIDIA went straight to the gaming anti-cheat software providers and got the popular competitive titles running natively on Arm (that was a big problem in the past). That, my friends, is a tactical call which is unglamorous, down in the weeds, and absolutely decisive. Strategy gets you onto the battlefield. Tactics are how you take the hill once you're standing on it. 🪖
And here's my favorite detail, the one that tells you everything about the operator running this company: he did all of that while simultaneously shoving Vera Rubin, NVIDIA's next-generation data-center platform, into full production, with OpenAI, Anthropic, xAI, Dell, Oracle, and CoreWeave already queued up as customers. A merely well-run company picks one front and fights it competently. Huang opened a second front, on the same stage, on the same day, and had the nerve to make it look effortless. Even the venue was a decision. Computex, in Taipei, in today's literal beating heart of the supply chain that makes every one of these chips possible, with every chipmaker and every reporter on Earth already crammed into the room. None of this was an accident. Every single piece was placed by hand.
When you see execution like this, you're watching something that almost never actually happens: a leader who can hold a ten-year vision in one hand and make the right snap decision with the other. History offers us a short list of these people. Steve Jobs is the obvious one, and the parallel here is almost too on-the-nose to be allowed–because the last time somebody walked away from Intel and won, it was Apple. Jobs set the course of owning the entire stack, and that bet eventually matured into the Arm-based M-series silicon that quietly buried Intel inside the Mac. He paired the long arc–the iPhone, controlling everything from sand to software–with ruthless tactical cuts: killing the floppy drive, refusing Flash, betting the company on a single button. Vision and the scalpel, in the same hands. Sidenote–watch Apple’s new CEO who is a hardware guy expected to leverage Apple’s proprietary silicon to provide AI on its products…just sayin’. 👀
Andy Grove belongs in this conversation too, and the irony is so rich it's almost cruel, because Grove built Intel. The man who literally wrote "Only the Paranoid Survive" yanked Intel out of the memory business and bet the whole company on microprocessors, a bet-the-farm strategic call executed with legendary tactical discipline. Intel today is getting comprehensively out-Groved by the company across the street, and somewhere the lesson is screaming into a pillow. Jeff Bezos ran the same playbook on a different field. "Day 1" (that’s Bezos’ well-known strategy) forever, years of thin margins swallowed for long-term position, and then the tactical masterstroke of turning Amazon's internal plumbing into AWS (first mover and leader in cloud computing), the side project that quietly became the profit engine of the whole empire. And Satya Nadella, who–and this is not a coincidence I'm going to let you ignore–will be standing right next to Huang later today at Microsoft’s Build 2026 conference. Nadella inherited a stagnant, widely-mocked Microsoft, bet the house on cloud, swallowed a generation of corporate pride to embrace open source, and made the OpenAI call that reshaped the company, and–to be fair–shot the starting gun for the AI race that we are deep into today. The CEO sharing Huang's stage is himself a walking case study in the exact skill we're talking about. That's the company Huang keeps. Not chip designers. Operators who saw around corners and then had the guts to act on what they saw.
Now I want you to do a little work, because this is not a buy-this, sell-that blogpost/newsletter–I don't tell you what to throw in your cart. Instead, go open your portfolio and read it back to yourself, name by name, and for every single company you own, ask the management question. Which of these CEOs is sitting on a CUDA-equivalent–a moat that took a decade to build and can't be copied next quarter? Which of them has made the patient, multi-year bet and then shown up with the tactical kill-shot the instant the window cracked open? Which of them, staring down a graveyard of failed attempts by everyone who tried before, would actually go fix the causes of death instead of shipping and hoping for the best? If you can answer that confidently for a holding, perfect, because that's a base worth standing on. If you can't–if you're white-knuckling a great product strapped to a management team you couldn't describe making one sharp decision under pressure–well, that's your homework for the week, and I'd get to it.
Let me keep myself honest, the way I always do, because one keynote announcement is not a coronation. NVIDIA’s RTX Spark could absolutely stall if it ships as a pricey luxury toy and nobody actually buys it (there is a graveyard full of those as well, 😉, IYKYK), and the analysts who do this for a living are right to keep waving that flag. Some analysts believe the PC market shrinks in 2026 on memory shortages and rising costs. The verdict the market handed down yesterday was sentiment, not a product that has shipped, sold, and won yet. But that, my friends, is precisely the point, isn't it? That's exactly why you bet on the management and not on the keynote. The part of this story that hasn't happened yet–the pricing, the supply crunch, the inevitable counterpunch from Intel and Qualcomm once they stop bleeding and start thinking– that uncertain, unwritten chapter is the exact rocky road great management is built to cross. It’s easy for a CEO to bask in the sunlight of a good launch. True operators earn their keep in the part you can't see coming. So enjoy the fireworks for now. Just don't go mistaking them for the foundation. Because in the end…it's all about the base.
YESTERDAY’S MARKETS
Yesterday, all three major indexes finished at record highs, with the Nasdaq Composite up 0.42%, the S&P 500 up 0.26%, and the Dow up 0.09%, driven by a technology rally following Nvidia's RTX Spark chip launch. 🙃 Breadth was narrow, as technology and energy were the only S&P 500 sectors to close higher while the VIX rose about 4.8% to 16.05. In commodities, Brent crude jumped nearly 7% to $97.37 a barrel after US airstrikes on Iranian radar and drone sites escalated the conflict, while gold fell roughly 2.1% to around $4,497 an ounce. Bitcoin slid about 3.6% to near $70,800.
NEXT UP
JOLTS Job Openings (April) is expected to come in unchanged at 6.87 million vacancies. Don’t sleep on this very important report–the Fed loves it, and so do I. 😉
Fed speakers today: Kashkari, Hammock, and Goolsbee.
Important earnings today: Dollar General, Victoria’s Secret, Palo Alto Networks, Gitlab, and Ulta Beauty.