NVIDIA Q2 Earnings Report: Behind $96.2 Billion in Revenue, AI Computing Is Turning from a “Story” into a “Business”

$96.2 billion in revenue, 106% year-over-year growth, and $89 billion in data center revenue—NVIDIA’s Q2 FY2027 earnings report has once again pushed the market’s imagination of what “AI chips” can mean to new limits. When a company’s quarterly revenue approaches $100 billion and net profit nears $60 billion, the question is no longer “Can it keep growing?” but “Where exactly is the ceiling?”

Data Center’s Absolute Dominance Is a Double-Edged Sword

Data center revenue accounted for 92.5% of total revenue, up 117% year over year. AWS has committed to deploying an additional 2 million GPUs, revenue from hyperscale customers rose 102% year over year, and AI cloud and enterprise customers grew 138%—these figures point to one thing: Global investment in AI infrastructure has not stopped; it is accelerating. But the 92.5% concentration also means NVIDIA’s fate is now deeply tied to the AI computing investment cycle. Once hyperscale cloud providers slow their capital expenditures, this extreme concentration will bring greater volatility than that faced by its peers.

Market Reaction: Beating Expectations Has Become the “New Normal”

After the earnings report was released, NVIDIA’s after-hours share price briefly climbed 4.7% to $219.53. Goldman Sachs raised its price target to $300, while Bernstein took a more aggressive approach and raised its target to $400. But unlike the previous pattern of an inevitable decline after earnings, the market responded positively this time—showing that investors are beginning to believe AI computing demand is not a bubble, but a long-term trend that is being realized.

Outlook: Supply, Not Demand, Is the Bottleneck

The company expects Q3 revenue of $108 billion and guided for approximately 70% revenue growth in FY2028. On the earnings call, management stated plainly that unconstrained real market demand could support 100% growth, with the only limiting factor being insufficient supply chain capacity. The shortage of computing supply is expected to continue through at least the end of FY2028—meaning that for the next two years, NVIDIA will still face the “sell everything it produces” kind of sweet problem.

Jensen Huang said, “Computing is revenue.” Judging from this earnings report, AI has evolved from a technological concept into a real business. For investors, the real test is not when demand will peak, but whether NVIDIA can continue breaking through bottlenecks on the supply side and maintain its lead with its next-generation products (Vera Rubin is already in full production). At least for now, this AI giant is still moving ahead at full speed.

#英伟达财报周