Nvidia reports second-quarter earnings on Aug. 26, and the headline numbers are already eye-catching: analyst consensus calls for revenue of $92 billion and adjusted earnings per share of $2.09. That would represent a 96 percent year-over-year revenue increase and continued acceleration from the prior quarter. The stock fell 2.9 percent to $208.48 on Monday, reflecting broader anxiety ahead of the report.
Data Center is the engine driving almost everything. Analysts expect that segment to generate $85.4 billion—up 107 percent year-over-year—making it the single largest and most watched line item in the release. The segment now breaks into two buckets under a reporting framework Nvidia introduced last quarter: Hyperscalers and AI Clouds projected at $43.5 billion, and Industrial and Enterprise (ACIE) projected at $41.7 billion.
The hyperscaler figure is where the real tension lives. Amazon, Microsoft and Google collectively account for more than half of Nvidia's total sales. Last quarter, just three customers represented over 50 percent of revenue. All three of those companies also run internal chip development programs designed to reduce their dependence on Nvidia's hardware. Amazon has its Trainium line, Google builds Tensor Processing Units, and Microsoft has invested in custom silicon alongside its Azure commitments. That creates a structural headwind no single strong quarter eliminates.
Microsoft, Amazon and Google all reported their own quarterly results nearly a month ago, each posting strong cloud growth that gave the AI trade a partial reprieve. Google and Meta separately flagged higher capital expenditure plans, which spooked investors watching for signs that AI infrastructure spending is outrunning the revenue it generates. The key question—whether hyperscaler customers eventually pull back on GPU orders as their own chips mature—is the overhang Nvidia's management team will need to address on the earnings call.
Nvidia's updated reporting framework is itself a signal worth watching. Breaking out Hyperscaler revenue from ACIE revenue gives investors a cleaner look at customer-concentration dynamics quarter over quarter. If hyperscaler revenue is tracking at $43.5 billion but ACIE surprises to the upside, that is evidence Nvidia's revenue base is broadening beyond its three largest customers. If ACIE disappoints while hyperscalers carry the number, the concentration risk story intensifies.
The PC, gaming, workstation, robotics and automotive businesses now sit inside Nvidia's Edge Computing segment under the new framework. That segment draws less attention from analysts fixated on data center, but it represents the part of Nvidia's business least exposed to the hyperscaler concentration question. Any upside there is a free option the market is largely ignoring.
Nvidia has been actively building outside its core chip sales. Earlier this month the company announced a partnership with BlackRock, Blackstone, KKR, Apollo, Brookfield and Goldman Sachs to create a $500 billion capital pool that would securitize Nvidia's GPUs—essentially turning its hardware into a financeable asset class. That structure, if it scales, reduces Nvidia's dependence on any single buyer making a large upfront purchase decision.
The company is also co-funding the construction of an 8-gigawatt data center in Ohio alongside SB Energy and OpenAI, with total investment projected at up to $150 billion. An 8-gigawatt facility would rank among the largest single data center projects announced globally. Nvidia's involvement on the capital side of that deal, rather than just the supply side, reflects a strategy to lock in long-term chip demand through equity stakes in the infrastructure that consumes its products.
Chip stocks more broadly have struggled over the past month. July brought steep declines across the sector as investors questioned whether AI capital spending would translate into measurable returns. Microsoft, Amazon and Google partially addressed those concerns through strong cloud revenue prints, but macro anxiety has not fully lifted. Nvidia's Aug. 26 report is the last major earnings datapoint for the AI infrastructure trade this cycle.
The numbers to watch: Does total Data Center revenue clear $85.4 billion? Does ACIE show accelerating growth independent of the three mega-cap hyperscalers? And what does management say about second-half demand given that hyperscaler customers are simultaneously ramping their own custom silicon programs? Those three questions will determine whether the stock's 2.9 percent drop looks like a buying opportunity or an early warning.
