Nvidia Earnings Top $96B as AI Buildout Accelerates

Nvidia reported quarterly revenue of $96.22 billion on August 27, more than double the same period a year earlier and roughly $4 billion ahead of the analyst consensus compiled by FactSet. Shares rose more than 8% during Thursday trading.

Founders have limited reason to care about a chipmaker’s quarter in isolation. This one is different, because the guidance embedded in it sets the budget assumptions for nearly every company selling software, and the wider AI spending picture now runs through these numbers.

Breaking Down the Quarter

Net income for the May to July period reached $59.69 billion, or $2.46 per share. A year earlier the comparable figures were $26.42 billion and $1.08 per share.

Adjusted earnings came in at $2.22 per share against a $2.09 consensus. Operating expenses climbed 55% to $8.41 billion, which is worth noting because it shows the company spending aggressively rather than coasting.

Data center sales, covering hyperscale cloud operators such as Amazon, Meta and Google, accounted for $89 billion of the total. Edge computing, which spans PCs, consoles and robotics, added $7.2 billion.

Guidance for the current quarter sits near $108 billion, above the $104.86 billion analysts expected. Hitting that figure would represent roughly 89% year-over-year growth, so the growth rate is still climbing rather than flattening.

Supply Is the Constraint, Not Demand

The most useful detail for founders was not the revenue line. It was CEO Jensen Huang telling analysts that the company’s entire supply chain is under strain and that available supply currently covers about 70% of what customers want.

CFO Colette Kress reinforced the point. She said the growth outlook for the fiscal year ending January 2028, currently around 70%, would be closer to double that figure if the company could source enough components.

Read that carefully. When a supplier is rationing, prices firm up and lead times stretch.

Any founder whose product depends on GPU capacity should therefore assume compute costs stay elevated rather than falling on the usual hardware curve. Kress also told analysts that central processing unit revenue should more than double in fiscal 2028, which signals Nvidia expanding into territory it did not previously dominate.

The Capex Wave Reaching Your Customers

Kress put hyperscaler capital spending at nearly $800 billion this year, rising toward $1.3 trillion in 2027. Money at that scale does not stay inside data centers. It flows outward into power, cooling, construction, networking and staffing.

That creates real demand for smaller companies positioned alongside the buildout. Meanwhile the consolidation trend continues, with deals like AI infrastructure consolidation reshaping which platforms founders end up building on.

The practical move is to check whether your buyer sits inside this spending wave or outside it. Companies selling into AI infrastructure are seeing budgets expand.

Everyone else is competing for budget that AI line items are steadily absorbing. That distinction should shape your pricing conversations this quarter, because a buyer with an expanding AI budget will hear a very different pitch than one defending flat spend.

The Bear Case Worth Holding Onto

Nvidia’s market value has climbed from about $400 billion at the end of 2022 to roughly $5.2 trillion. Concentration on that scale makes a lot of investors uneasy, and the skepticism is not unreasonable.

Two specific risks stand out. The outlook assumes no data center compute revenue from China at all. Local opposition to data center expansion is also growing, which could slow the physical buildout regardless of chip availability.

A third concern sits underneath both. Analysts have questioned for two years whether AI applications will generate returns matching the trillions committed to building them, and this quarter does not settle that argument.

Huang described the shift bluntly, saying “compute is revenue” now. That framing works while customers keep converting compute into profitable output. It gets tested the moment they cannot.

Signals to Track Into September

Watch hyperscaler capital expenditure guidance in the next earnings round. If Amazon, Microsoft, Google or Meta trims plans, the downstream effect reaches suppliers within two quarters.

Also watch component pricing. Nvidia flagged rising memory costs pressuring its 75% gross margin next quarter, and that pressure travels through the hardware chain toward anyone buying servers or dedicated capacity.

For founders raising capital, the read is straightforward. Investor appetite for AI-adjacent businesses remains strong, though the bar has moved toward evidence. Recent moves in venture capital funding point toward the application layer rather than infrastructure.

Does this quarter mean AI demand is permanent?

No. It confirms that spending commitments extend through 2027, which is a planning horizon rather than a guarantee.

Should founders expect compute costs to fall?

Not soon. With supply covering roughly 70% of demand, rationing tends to hold pricing firm well into next year.

Does this change how I should price my product?

Possibly. If your costs track GPU capacity, build a margin buffer now rather than absorbing increases later.

Where can I verify these figures?

Nvidia posts full results and the analyst call materials on its investor relations site.

The bottom line for anyone building a company: the AI buildout has a funded runway through 2027, and supply rather than demand is currently setting the pace. Plan your costs around scarcity, and plan your sales around which side of the spending wave your buyer sits on.