Nvidia Earnings Reinforce AI Demand With $80 Billion Buyback

Nvidia earnings again set the pace for global technology markets on May 20 after the chipmaker reported record quarterly revenue, expanded its share repurchase plan, and projected another strong period for AI infrastructure spending. The results reinforced Nvidia’s central role in the buildout of data centers that power large language models, AI agents, and other compute-intensive services.

The company said revenue for the quarter ended April 26 reached $81.6 billion, up 85% from a year earlier and 20% from the prior quarter. Nvidia also announced an additional $80 billion share repurchase authorization and raised its quarterly cash dividend to 25 cents per share from 1 cent, underscoring how much cash the business is producing even as it keeps investing aggressively.

Nvidia Earnings Put Scale and Cash Generation in Focus

Nvidia’s latest quarter mattered well beyond one company because investors increasingly treat the chipmaker as a proxy for the durability of the broader AI spending cycle. When Nvidia delivers numbers above expectations, it tends to validate demand assumptions across semiconductor supply chains, cloud infrastructure providers, and software companies building on top of that hardware base.

This set of results did exactly that, while also showing that the market has become harder to impress. Reuters reported that Nvidia forecast second-quarter revenue of $91 billion, plus or minus 2%, ahead of Wall Street expectations, yet the shares still fell in extended trading as investors weighed whether the next phase of growth can remain as extraordinary as the last one.

Nvidia Earnings Beat Expectations Again

According to Nvidia’s earnings release, quarterly revenue rose to $81.6 billion from $44.1 billion a year earlier. GAAP diluted earnings per share climbed to $2.39 from $0.76, while operating income increased 147% year over year to $53.5 billion, illustrating how much of the revenue growth is still dropping through to profit.

AP reported that analysts polled by FactSet had expected revenue of $78.91 billion and earnings of $1.75 per share, meaning Nvidia once again cleared a very high bar. That matters because Nvidia has spent the past three years becoming the benchmark that shapes investor expectations for nearly every company exposed to AI demand.

The company also said it returned roughly $20 billion to shareholders during the quarter through share repurchases and dividends. Together with the new authorization, that capital return policy gives management another way to signal confidence in future cash flows even as spending on research, networking, software, and product transitions remains heavy.

Buyback and Dividend Change the Conversation

The additional $80 billion buyback was one of the clearest market-moving elements in the report. For a company already carrying enormous expectations, a buyback of that size suggests management believes the business can support both major internal investment and significant shareholder returns at the same time.

Nvidia also lifted its quarterly dividend to 25 cents per share from 1 cent. The dividend increase is small relative to the company’s valuation, but symbolically it broadens the message of the quarter: Nvidia is no longer being judged only as a fast-growing AI supplier, but also as a company with the balance sheet strength and cash generation profile of a mature market heavyweight.

Even so, the modest after-hours share reaction showed that investors are now asking a different question. The issue is no longer whether Nvidia can produce another big quarter, but whether it can keep stretching the AI investment cycle far enough into 2027 and 2028 to justify the standards the market has already built into the stock.

AI Infrastructure Remains the Core Nvidia Earnings Story

The most important line in the report was still the demand coming from data centers. Nvidia said data center revenue reached a record $75.2 billion, up 92% from a year earlier, showing that hyperscalers, AI-native cloud providers, and enterprise-focused AI infrastructure operators are still buying at a scale few industrial buildouts have matched.

That supports the broader argument that AI spending is shifting from experimentation toward installed capacity. Instead of treating AI as a discretionary software project, large customers are continuing to fund physical infrastructure, networking, accelerated computing, and model-serving systems as a long-cycle strategic requirement.

Data Center Growth Keeps Nvidia Earnings Dominant

By Nvidia’s own breakdown, data center compute revenue rose to $60.4 billion and networking revenue climbed to $14.8 billion. Those figures matter because they show that the opportunity is not just in graphics processors themselves, but in the wider architecture needed to make AI factories work at scale.

Reuters said Huang told analysts that Nvidia believes it can grow faster than hyperscale capital expenditure because a newer class of customers is emerging. He pointed to AI-specific cloud companies as a meaningful sub-segment, saying those customers are now contributing revenue at roughly the same level as the largest cloud operators while expanding faster quarter over quarter.

That is an important shift for the wider market. If demand is broadening beyond Alphabet, Amazon, and Microsoft, the AI infrastructure cycle may prove more resilient than critics expected, because the customer base becomes less concentrated and more tied to national, industrial, and enterprise buildouts.

Nvidia Earnings Also Show a Changing Product Map

Nvidia used the quarter to underline that its future growth is not confined to one chip family. The company introduced a new reporting framework built around Data Center and Edge Computing, while highlighting new Vera processors, networking products, inference software, and enterprise agent tools during the quarter.

The reporting change is not just cosmetic. It suggests Nvidia wants investors to evaluate it less as a conventional semiconductor vendor and more as a platform company whose addressable market spans cloud infrastructure, AI factories, enterprise deployment, networking, industrial systems, robotics, and edge computing.

Management also said its second-quarter guidance does not assume any data center compute revenue from China. That detail is notable because it shows Nvidia still sees enough demand elsewhere to forecast $91 billion in quarterly revenue even without contribution from a market that has been strategically important but politically constrained.

Nvidia Faces Higher Expectations, Not a Simpler Market

Strong numbers do not remove the strategic pressures around Nvidia. In some respects, they intensify them, because every quarter of exceptional growth raises the threshold for what counts as good enough. Investors now want proof not only of near-term demand but also of how durable Nvidia’s pricing power and technology edge will be as competitors respond.

That tension was visible in the market reaction. AP reported that the stock dipped slightly after hours despite the beat, while Reuters described investors as focusing on whether inference workloads, custom silicon, and broader competition could eventually make growth less explosive than it has been during the first phase of the generative AI boom.

Competition Is Becoming Part of the Nvidia Earnings Debate

Many of Nvidia’s biggest customers remain dependent on its systems, but they are also designing their own chips and software stacks. Reuters noted that Google, Amazon, AMD, and Intel are all trying to capture a larger portion of AI workloads, especially as customers look for alternatives in inference, cost control, and specialized deployment.

That does not mean Nvidia’s position is weakening today. On the contrary, the current quarter shows its products remain the standard around which much of the ecosystem is still organized. However, the market is beginning to distinguish between near-term dominance and the harder question of how much of the next computing stack Nvidia will ultimately own.

The company’s response is to keep widening the scope of what it sells. By tying chips to networking, software, inference optimization, and full-system design, Nvidia is trying to make itself harder to displace even if some customers continue developing internal alternatives.

What Nvidia Earnings Mean for the Rest of the Market

Nvidia’s results matter because they serve as a read-through for suppliers, cloud platforms, capital spending plans, and public-market sentiment around AI. When the company reports growth this strong, it tends to support valuations across semiconductor equipment makers, packaging specialists, server manufacturers, optics providers, and infrastructure developers.

Reuters also pointed to expectations that major U.S. technology companies could spend more than $700 billion on AI infrastructure this year, up sharply from about $400 billion in 2025. If those spending plans hold, Nvidia will remain one of the clearest channels through which investors measure how quickly AI enthusiasm is turning into physical economic buildout.

At the same time, the restrained stock reaction is a reminder that markets have moved from amazement to scrutiny. Nvidia no longer needs to prove AI demand exists. It needs to prove that demand can keep compounding at a pace large enough to satisfy investors who already treat extraordinary performance as the starting point rather than the upside case.

Nvidia earnings strengthened the case that the AI infrastructure cycle still has momentum, but they also showed how much future belief is already embedded in the market’s expectations. For more context on the companies, policy shifts, and capital flows shaping that race, continue reading related coverage at Berrit Media.