Most analysts anchoring Nvidia’s path to a $10 trillion market cap treat it as a revenue story. Get to roughly $600 billion annually at a 50% net margin, apply a reasonable earnings multiple, and the math closes. That framing is correct but incomplete. The more interesting question is what changed structurally in August 2026 that makes $600 billion look conservative rather than ambitious.
The Per-Gigawatt Argument
During Nvidia’s second-quarter fiscal 2027 earnings call, CEO Jensen Huang explained that the company’s revenue opportunity has climbed from roughly $18 billion per gigawatt with Hopper to $25 billion with Grace Blackwell and now $40 billion with Vera Rubin. That single data point reframes the entire debate. AI infrastructure spending is constrained by power availability, not by customer willingness to spend. As AI data centers become constrained by electricity, performance per watt becomes an increasingly valuable competitive advantage, and Nvidia’s ability to raise its revenue opportunity from $18 billion per gigawatt with Hopper to $40 billion with Vera Rubin suggests its moat is not merely about making the fastest GPU.
The company began production shipments of Vera Rubin in fiscal Q3 2027, and the ramp is underway.
The Revenue Base Is Already Extraordinary
Nvidia reported revenue of $96.2 billion for the second quarter ended July 26, 2026, up 106% from a year ago, with data center revenue of $89.0 billion, up 117%. The annualized run rate at that pace exceeds $385 billion. Management also discussed a goal of about 70% growth in fiscal 2028, which would push full-year revenue toward $650 billion on the current trajectory, comfortably past the threshold most models require for a $10 trillion valuation.
Sovereign AI revenue more than tripled to over $30 billion in fiscal 2026, and that demand line is structurally different from hyperscaler spending. Governments buying AI infrastructure treat it as a national utility, not a discretionary capex cycle. Management has framed national AI infrastructure investment as analogous to utility build-out. That analogy matters: utilities don’t cancel orders when earnings disappoint.
The demand is structural on the buyer side too. Microsoft has committed to scaling its data center network toward 38 gigawatts of total power capacity by 2032, a buildout that locks in multi-year GPU procurement regardless of any single earnings cycle.
The Options Angle
Nvidia’s next earnings report is confirmed for November 17, 2026, after the market closes, with the stock near a market value of about $5.16 trillion. Options flow was active in late October, but contract-volume and put/call figures vary materially by data vendor and calculation method, so treat any single-day ratio with caution.
A January 2028 call spread targeting the $280 to $310 range captures the fiscal 2028 revenue ramp with defined risk, avoiding the full premium cost of outright LEAPS into a name carrying elevated implied volatility. Management has noted supply constraints are limiting growth with the supply chain running at full capacity, and gross margin is expected to moderate from 75% in fiscal Q2 toward 71-72% in fiscal Q4. Margin compression is the primary thesis risk to monitor heading into the November report.
The Beast Verdict
The $10 trillion argument was theoretical six months ago. Some scenario work has put Nvidia’s potential $10 trillion milestone around late 2030, with a wider range of 2029 to 2032. The Vera Rubin per-gigawatt dynamic compresses that range. When a company earns 60% more revenue from the same physical power footprint, the ceiling rises without requiring a single new data center to be built. That is not a growth story. That is a pricing power story, and it is the most underappreciated line in Nvidia’s entire bull case.
