NVIDIA Posts Record $96.2B in Q2 Revenue, Guides to $108B for Q3

$96.2 billion. That is what NVIDIA printed in quarterly revenue for Q2 2026, doubling its result from the same period a year earlier and setting a new all-time high for the chipmaker as demand for AI computing infrastructure shows no signs of slowing.

The Santa Clara company reported 75% gross margins alongside the revenue record, a figure that underscores how thoroughly NVIDIA has priced its position at the center of the global AI buildout. Data center remained the dominant growth engine, as it has been since generative AI demand reshaped the company's business mix starting in 2023, when gaming revenues were eclipsed by enterprise compute orders.

Forward guidance came in at $108 billion for Q3, above analyst consensus. The company disclosed $366 billion in future customer commitments, signaling that hyperscalers, sovereign AI programs, and enterprise buyers have locked in purchases well beyond the current quarter. NVIDIA also disclosed up to $108.5 billion in guarantee exposure, a contingent liability figure reflecting the scale of long-term supply agreements the company has entered to meet that demand. NVDA shares jumped in after-hours trading on Wednesday following the announcement.

The results cement a trajectory that began accelerating in late 2022 with the public release of large language models and the subsequent race by cloud providers to build out GPU (graphics processing unit) clusters at scale. NVIDIA's H100 and Blackwell-architecture chips have become the de facto standard for training and inference workloads, giving the company pricing power that its 75% gross margin reflects directly. For context, semiconductor peers typically operate gross margins in the 50-60% range; NVIDIA's figure is exceptional even by its own historical standards.

Not every signal is unambiguously positive. The $366 billion in commitments creates execution risk: if AI capital expenditure cycles turn or customer priorities shift, unfulfilled orders could reverse. Competition is intensifying from AMD's MI-series accelerators, Intel's Gaudi line, and custom silicon built in-house by Google, Amazon, and Microsoft. Export restrictions on advanced chips to China also remain a structural overhang on international revenue. At current valuations, NVIDIA's stock already embeds years of strong growth, leaving limited margin for disappointment.

For crypto and blockchain infrastructure, NVIDIA's results carry indirect weight. GPU compute underpins a range of decentralized AI and distributed computing networks. Sustained demand from hyperscalers keeps GPU supply tight and prices elevated for smaller buyers, including decentralized physical infrastructure networks (DePIN) that rely on commodity GPU availability. A prolonged AI infrastructure supercycle, which NVIDIA's $108 billion Q3 guide implies, keeps that supply constraint in place.

NVIDIA also announced expanded AI infrastructure partnerships in the days following the earnings release, though specific terms were not disclosed in the company's announcement. The pattern of post-earnings partnership disclosures has become a regular feature of NVIDIA's communications cadence, reinforcing its role as the connective tissue across the AI supply chain.

The Q3 guide of $108 billion, if achieved, would represent another sequential record and push NVIDIA's annualized revenue run rate above $430 billion. At that scale, the company's results function less as a single-stock story and more as a real-time indicator of where global AI capital is flowing.