“Nvidia hits an all-time high—AI is back!” has been the market’s headline for the past 24 hours, and I can see why. As of this morning, shares are trading around $183–$184 after tagging a record intraday high near $184.5; the tape is acting like the AI trade just got fresh oxygen. The immediate spark isn’t hard to find, Nvidia said it plans to invest up to $100 billion in OpenAI in a sweeping partnership tied to building out massive AI data-center capacity over the next few years. That news sent the stock to fresh highs and re-centered the narrative around Nvidia as the indispensable arms merchant of AI infrastructure.
I don’t base my view on headlines alone. The most important anchor for me remains the company’s Q2 fiscal 2026 results. Revenue came in at $46.7 billion—up 6% sequentially and 56% year over year—with Data Center revenue at $41.1 billion. Management guided Q3 FY26 revenue to about $54 billion (±2%), and paired that with mid-70% non-GAAP gross-margin commentary later in the year. Those are extraordinary figures for a company this large, and they reaffirm my core thesis: Nvidia is still the central supplier for the training and inference build-out happening across hyperscalers, model labs, sovereign projects, and now a widening set of enterprise workloads.
What I pay equal attention to are the risks that the market is now starting to price with more nuance. The press release explicitly noted no H20 shipments to China in the quarter and said the Q3 outlook assumes none—even though the company has received some export licenses. Management also detailed a $180 million release of previously reserved H20 inventory and about $650 million of H20 sales to a non-China customer. In plain English: China remains a swing factor, and while the rest of the world is absorbing a tremendous amount of Nvidia compute, the company’s most contentious geography is not in the near-term numbers. That’s a material macro variable I can’t ignore.
Capital returns add a second pillar to my view. Alongside Q2 results, Nvidia disclosed that it had returned $24.3 billion to shareholders in the first half of FY26 and that its board authorized an additional $60 billion for share repurchases without expiration. For me, that matters for two reasons. First, it signals management’s confidence in multi-year cash generation tied to Blackwell-class systems and the software stack around them. Second, it gives the company a powerful tool to dampen volatility and support per-share earnings if macro headlines shake the tape. In a market that can obsess over quarter-to-quarter order cadence, that much buyback firepower is a real backstop.
NVDA Daily Chart
At today’s prices, $200 is only a short step above spot, not a deep-discount entry. If I’m thinking like a long-term owner, I can justify adding here precisely because the operational backdrop (Q2 execution, Q3 guide, buyback, OpenAI alignment) is supporting the move to new highs rather than chasing on fumes. If I’m thinking like a tactician, I’m not compelled to wait for exactly $200; my preference is to scale entries—a starter tranche near current prices, then add on either a clean post-news consolidation or a breakout with volume. The reason is simple: the all-time high often invites volatility as fast money takes profits, and I’d rather average into strength or buy the retest than bet on a precise print. With China still an overhang and expectations sky-high, disciplined sizing matters more to me than landing a perfect round number.
The OpenAI partnership talk is also worth parsing. The announcement was framed as a massive, staged investment and infrastructure plan, with OpenAI deploying at least 10 gigawatts of Nvidia-powered compute over time and Nvidia taking a non-controlling equity stake. The structure and exact phasing still need to be finalized, but the strategic read-through is clear to me: this tightens Nvidia’s position at the very center of the frontier-model ecosystem and sets an explicit multi-year compute roadmap that aligns with Nvidia’s next-gen platforms. In a world where customers are weighing custom silicon and alternative supply chains, the strongest answer is continued proof that the most demanding model developers still choose Nvidia for performance at scale.
I’m not blind to where this can bite. When one company powers so much of the AI stack, customer concentration becomes a double-edged sword—if a megacap cloud buyer pauses or if geopolitics crimp a region, the quarterly growth slope can wobble. Valuation also isn’t trivial; at fresh highs, the market is effectively underwriting years of elevated demand, healthy margins, and continued ecosystem lock-in. That’s why my playbook is to own, but size rationally—I want exposure commensurate with Nvidia’s centrality to AI, while keeping dry powder for the inevitable headlines that test conviction.
Yes, the AI narrative is back—but more importantly, the fundamentals never left. Nvidia just printed a record quarter, guided higher, armed itself with a colossal buyback, and moved to deepen ties with the most visible AI lab on the planet—all while openly acknowledging the China risk and excluding it from guidance. For me, that transparency plus execution is exactly the mix I want to see at all-time highs. If I’m building a position, I’m comfortable initiating here and layering in methodically rather than waiting for a perfect $200 tag. If the stock gives me a calmer consolidation or a post-breakout retest, I’ll lean in; if it rips, I’ll let the starter sit and look for the next orderly entry. Either way, the center of gravity for AI infrastructure is still Nvidia, and that’s where I want some of my capital to live.
@MillionaireTiger @Tiger_comments @Daily_Discussion @CaptainTiger @TigerSG
Disclaimer: This is a general analysis and not financial advice. Views in this post may be biased. Always conduct your own research before making any investment decisions.
Comments