Nvidia earnings preview: AI rally faces its next reality check

Key points

• Nvidia’s earnings are likely to be the next major test for the AI trade. The company remains at the centre of the AI infrastructure boom, but expectations are high after a strong rally in technology and semiconductor stocks.
• The market may focus more on guidance than the past quarter. Investors will want to know whether AI demand is still accelerating, whether margins remain strong and whether China or competition risks are becoming more material.
• The key risk is that “good” may not be good enough. Nvidia has become a signal for the broader AI theme, so any disappointment could spill over into semiconductors, Nasdaq and other AI-linked names.

Nvidia is expected to report fiscal Q1 2027 earnings for the quarter ended 26 April 2026 on 20 May 2026, and this is no longer just a company-specific event. For many investors, Nvidia has become the scoreboard for the AI boom.

The company sits at the centre of cloud computing, data centres, generative AI and high-performance chips. Its results help answer one big market question: is AI infrastructure spending still running hot, or is the market starting to price in too much good news?

That question matters more now because technology and semiconductor stocks have already rallied strongly. The AI story is still powerful, but the short-term setup is less one-sided. Nvidia may need to deliver not just strong results, but strong enough results to justify the recent move.

The numbers to watch — And why they matter

Based on Bloomberg consensus estimates for the quarter ending April 2026, investors may want to focus on a short list of numbers rather than every line item.

1. EPS and revenue: Can growth still surprise?

Consensus is looking for adjusted EPS of around $1.77 and revenue of around $78.9 billion for the quarter.

That is already a very high bar. Nvidia has trained the market to expect strong earnings beats, so the key question is not simply whether it beats. It is whether the beat is big enough, and whether guidance suggests the growth story is still accelerating.

2. Data centre and networking: the heart of the AI story

Data centre revenue is expected at around $73.3 billion, or roughly 87% year-on-year growth.

This is the most important number because it captures demand from cloud providers, AI labs and companies building AI infrastructure. Investors will watch whether customers are still racing to add capacity, and whether Blackwell demand is ramping smoothly.

Rubin will also matter for the forward story. Investors will want confidence that Nvidia has another major platform transition ahead, rather than a one-off Blackwell upgrade cycle.

Networking is becoming an important part of this story too. AI data centres do not just need more chips; they need faster connections between chips, servers and clusters. That is why Nvidia’s networking business is being watched for record demand, supported by larger AI clusters, Ethernet/InfiniBand adoption and the need to reduce bottlenecks in training and inference workloads.

3. Margins and supply: Is pricing power still intact?

Adjusted gross margin is expected around 75%.

Margins matter because they show whether Nvidia still has strong pricing power. High margins would suggest demand remains strong, supply remains tight and Nvidia’s platform advantage is intact.

Supply is worth watching because demand for Nvidia’s newest AI systems has been stronger than the industry can easily absorb. The question is not only whether customers want more Blackwell chips, but whether Nvidia and its partners can deliver enough complete systems — including advanced packaging, high-bandwidth memory, servers and networking equipment — on time. If supply improves, Nvidia can convert more demand into revenue. But if bottlenecks persist, some sales may be pushed into later quarters, while faster capacity expansion could also test margins if costs rise.

4. AI capex: Is the spending cycle still expanding?

Nvidia’s revenue shows what customers are buying now, but hyperscaler capex plans show whether the AI buildout still has legs.

Investors will listen for signs that large cloud and technology companies are continuing to spend heavily on GPUs, networking and full-stack data centre infrastructure. This matters because Nvidia’s growth depends not only on demand for chips, but on the scale of the entire AI factory buildout.

The positive read would be that customers are still expanding AI capacity and need more compute, networking and software to support larger models and inference demand. The negative read would be any sign that customers are becoming more selective, slowing orders or facing investor pressure to prove returns on AI spending.

5. Guidance: The real market mover

The most important part of the report may be what Nvidia says about the next quarter and the broader demand outlook.

This is where the bar is already very high. Bloomberg consensus currently points to Nvidia revenue rising from about $216 billion in FY2026 to around $370 billion in FY2027, implying roughly 71% annual growth. Adjusted EPS is also expected to rise from about $4.77 to $8.43, or around 77% growth.

That means investors are not pricing in a normal growth year. They are pricing in another step-change in AI demand, supported by Blackwell, early Rubin visibility, networking strength and continued hyperscaler capex.

A strong guide would suggest the AI infrastructure cycle still has momentum and that the market’s high growth assumptions remain achievable. A cautious guide, even after a solid quarter, could trigger profit-taking because the stock has already rallied sharply and expectations for FY2027 are demanding.

For investors, this is the key lesson: a good company can still see its share price fall if expectations were too high going into the results. Nvidia may not only need to beat today’s numbers; it may need to defend tomorrow’s growth story.

Positive drivers to watch

The most constructive outcome would be a combination of strong revenue, strong data centre sales, resilient margins and upbeat guidance.

That would suggest AI demand is still broadening, Nvidia remains the key supplier of AI infrastructure and customers are still willing to spend aggressively despite higher costs and macro uncertainty.

Other positive drivers could include:

• Management commentary that demand remains strong across cloud providers and enterprise customers.
• Evidence that supply constraints are easing without hurting pricing power.
• Confidence that new products are supporting another leg of growth.
• Signs that networking and full-stack AI infrastructure demand remain strong, not just chip demand.
• Continued confidence from hyperscalers that AI capex will remain elevated.

If these drivers come together, the report could support the broader AI trade and improve sentiment across semiconductors, AI infrastructure names and the Nasdaq.

Negative drivers to watch

The main risk is not necessarily a weak quarter. The bigger risk is a quarter that is good but not good enough.

Negative drivers could include:

• Guidance that is only in line with expectations or sounds more cautious.
• Any sign that data centre demand is slowing or becoming more selective.
• Margin pressure from product transitions, higher costs or customer pricing pressure.
• Supply constraints that delay deliveries, limit near-term revenue conversion or make the Blackwell/Rubin transition look less smooth than expected.
• More uncertainty around China demand, export restrictions or local procurement policies.
• Rising competition from AMD or custom chips developed by large cloud companies.
• Any concern that hyperscalers are slowing AI capex, becoming more selective, or facing investor pressure to show returns on AI spending.

Valuation and earnings history: The bar is high, but not impossible

Nvidia’s earnings history shows why investors still give the stock the benefit of the doubt. Over the past several quarters, the company has often beaten consensus expectations, with Bloomberg data showing an average earnings surprise of around 15%. That track record is one reason markets continue to expect upside risk around results.

But the share-price reaction has become less automatic. Recent results have shown that even earnings beats can be followed by profit-taking if the guidance is not strong enough or if investors had already priced in a very strong quarter. In simple terms: Nvidia has trained the market to expect beats, so the surprise now needs to be more meaningful.

Valuation adds another layer. Bloomberg data shows Nvidia trading at roughly 48x trailing earnings currently, lower than the extreme multiples seen earlier in the AI cycle but still a premium valuation that depends on continued strong growth. Consensus expects earnings growth to stay very high into FY2027, so the stock may look more reasonable if Nvidia keeps delivering — but more vulnerable if growth expectations are revised lower.

The takeaway for investors: valuation is not the main problem if the AI growth cycle keeps accelerating. The risk is that a premium stock with premium expectations has less room for disappointment.

Bottom line

Nvidia’s earnings are likely to be a reality check for the AI rally.

The long-term AI infrastructure theme remains powerful, but the near-term bar is high.

The company does not only need to show that demand remains strong. It may need to show that demand is still accelerating, margins remain resilient, competition risk is contained, supply constraints are manageable and China risk is manageable.

That is a high bar. And when the bar is high, even good news can create volatility.