The market is no longer fixated on the beat alone
Nvidia’s fiscal 2027 second-quarter results showed that AI infrastructure demand remains strong, but the discussion after earnings has clearly shifted. The biggest questions are no longer limited to another revenue beat. Investors are now watching Rubin’s production ramp, pressure on gross margin, and a forward outlook that excludes China data center compute revenue.
Revenue for the quarter came in at $96.2 billion, up 106% from a year earlier. Data center revenue reached $89 billion, up 117%. Nvidia’s official fiscal 2027 second-quarter release also signaled that the company has moved into its next platform cycle, with Vera Rubin now in full production while AI infrastructure spending continues to expand.
At Nvidia’s current scale, that alone is not what the market is parsing most closely. The real focus has shifted to whether Rubin can carry the growth cycle forward and whether Nvidia can defend profitability as memory and system costs rise and China remains constrained.
Key figures from the quarter
- Fiscal 2027 second-quarter revenue: $96.2 billion.
- Data center revenue: $89 billion, up 117% year over year.
- Vera Rubin has entered full production.
- Fiscal 2027 third-quarter revenue outlook: about $108 billion, plus or minus 2%.
- That outlook excludes China data center compute revenue.
- Non-GAAP gross margin is expected to move from 75.0% in the second quarter to about 74.0% in the third quarter.
Strong results, but a higher bar
The quarter itself offered little evidence of a broad slowdown in AI infrastructure demand. Revenue rose 18% sequentially to $96.2 billion. Data center revenue also increased 18% from the prior quarter to $89 billion. Non-GAAP diluted earnings per share came in at $2.22.
Still, the post-earnings reaction suggested that investors are judging Nvidia differently than they did earlier in the AI cycle. The stock initially moved lower, then reversed as management’s conference call progressed. Attention turned quickly to future AI demand and the pace of Rubin production. Reuters, in its report on Nvidia’s post-earnings stock reaction, said the early hesitation reflected how high expectations had already become, rather than any clear deterioration in AI demand.
That distinction matters. Earlier in the cycle, a large beat by itself could lift expectations all over again. Now the market is asking whether each new Nvidia platform can trigger another round of infrastructure spending. In that setting, forward indicators matter more than the quarter that has already been reported.
Rubin has become Nvidia’s main growth test
The most important line in the report may not have been a revenue figure at all. It was the statement that Vera Rubin has entered full production.
Nvidia said partners including CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius have already begun running Rubin systems. That points to a rapid transition from the Blackwell deployment cycle into the next major architecture shift.
The speed of that transition matters because Blackwell has already shown that hyperscalers, AI labs and cloud providers are willing to keep deploying larger accelerated computing systems. Rubin now has to show that this spending carries into the next generation of hardware rather than ending as a one-off wave of capital expenditure.
Reuters said in its analysis of Nvidia’s post-earnings outlook that Rubin may already account for about 20% of Nvidia’s data center revenue this quarter. If that mix develops as expected, Rubin is no longer just a roadmap product. It is already becoming part of near-term revenue growth.
That leaves investors watching several things at once: the speed of Rubin deployments, the company’s ability to supply full systems, customer adoption patterns, and whether cloud providers remain willing to expand AI capacity fast enough to absorb each new Nvidia hardware cycle.
AI demand remains strong, but the debate has changed
Nvidia expects fiscal 2027 third-quarter revenue of about $108 billion, plus or minus 2%, extending its rapid quarterly growth trend.
The current numbers still do not point to a broad slowdown in AI infrastructure demand. Nvidia said demand is expanding beyond frontier AI labs and cloud providers into enterprises, sovereign AI projects and physical AI applications.
But the market’s argument is now more demanding. The central issue is no longer whether customers still need more GPUs. It is whether this level of AI infrastructure spending can generate enough economic value to justify the next round of capital deployment.
That helps explain why Nvidia’s product roadmap matters so much. If Rubin improves the economics of training and inference in a meaningful way, customers have a reason to keep upgrading and adding capacity. If returns on AI infrastructure become harder to prove, demand can remain large while spending turns more selective.
Based on the second-quarter results, the evidence still leans toward the first case. Infrastructure buildouts remain strong, and the next-generation platform has already entered production.
Margins may matter more than another beat
Profitability is likely the most important secondary signal in this report.
Nvidia posted a non-GAAP gross margin of 75.0% in the second quarter. The company expects that figure to fall to about 74.0% in the third quarter, plus or minus 50 basis points. A one-point drop alone does not imply a serious breakdown in fundamentals. What matters more is the business mix behind it: Nvidia is moving from selling GPUs into delivering more complex, full AI systems.
Modern AI infrastructure requires HBM and other memory, high-speed networking gear, advanced packaging, optical interconnects, power and cooling systems. When demand rises across all of those components at the same time, supply constraints and higher costs can affect the profitability of full-system sales.
Reuters reported that Nvidia expects higher memory and component costs to keep weighing on margins, and that gross margin could even slip to about 71%–72% in the fourth quarter. The same report examined the tension between accelerating Rubin demand and rising system costs.
That makes margin a metric investors will keep tracking. Nvidia has already shown that it can deliver exceptional revenue growth. The next question is how much of that growth it can convert into profit as each generation of AI infrastructure becomes larger and more dependent on a broad mix of components.
If Nvidia’s pricing power offsets those system cost increases, the pressure may stay manageable. If component inflation rises faster than the company can pass costs through to customers, margin trends will carry more weight in the NVDA investment case.
China is now framed as upside, not the base case
China remains strategically important, but Nvidia’s latest guidance changes how the market is treating that variable.
The company said clearly that its fiscal 2027 third-quarter revenue outlook does not include China data center compute revenue. In other words, the roughly $108 billion forecast does not rely on a near-term recovery in China AI accelerator sales.
That does not make China irrelevant. If US export policy changes, or if Nvidia can offer products that meet the rules, China could still represent meaningful incremental revenue.
For now, though, the structure of expectations has changed. Ongoing restrictions are already embedded in the near-term baseline. Any future reopening of market access would represent upside on top of that baseline. In the context of this earnings report, Rubin execution and global AI infrastructure demand appear more important than China revenue itself.
What the report signals for the wider AI semiconductor chain
Nvidia remains at the center of the AI computing ecosystem, and its earnings increasingly reflect what is happening across adjacent markets.
New Nvidia accelerator deployments do not expand in isolation. Large AI clusters also need more HBM and other memory capacity, faster networking, optical interconnects, advanced packaging, power infrastructure and additional data center capacity.
Nvidia also highlighted its Spectrum-6 networking system this quarter. The system supports both pluggable optics and co-packaged optics and will form part of the Rubin platform. That detail adds to the case that the next AI hardware cycle looks more like a full infrastructure buildout than a GPU-only story.
For semiconductor investors, the read-through goes beyond whether NVDA shares rise or fall right after earnings. If Rubin keeps deploying quickly, demand may remain strong across several AI infrastructure bottlenecks. At the same time, rising component costs suggest that companies will not benefit in the same way. Some suppliers may gain from scarcity and pricing power, while downstream system cost pressure increases.
That is why Nvidia’s earnings still matter for memory, networking, optical communications and data center names even when those companies are not direct GPU competitors.
The source article also noted that users tracking cross-equity moves can view Nvidia and other US stocks through MEXC’s US equities market section.
Four forward indicators for NVDA after earnings
With the quarter now reported, four signals stand out as more important than another discussion about whether Nvidia beat expectations.
- Rubin execution: the market wants confirmation that the transition from Blackwell to Rubin is proceeding smoothly, without major deployment or supply issues, and that customers remain willing to fund the new architecture.
- Gross margin: the move from 75.0% in the second quarter to about 74.0% in the third quarter, along with the possibility of heavier cost pressure toward year-end, puts memory prices, product mix and Nvidia’s pricing power in sharper focus.
- Hyperscaler and AI lab capex: Nvidia’s growth path depends on large customers continuing to build AI infrastructure at very high scale.
- China as optional upside: because third-quarter guidance already assumes zero China data center compute revenue, any policy shift that expands market access would change the baseline rather than simply preserve it.
Put together, Nvidia’s next challenge is no longer to prove that AI demand exists. It is to show that the demand can carry into the next architecture cycle while preserving the unusually high profitability built during the Blackwell era.
MEXC-related product references in the source article
The source article said Nvidia’s earnings may affect not only NVDA price action but also sentiment across the wider AI semiconductor and infrastructure supply chain.
Eligible users can trade real US-listed stocks through MEXC RealStocks, with services provided by regulated brokerage partners, according to the source. Product availability depends on a user’s region and qualification requirements.
For users focused on short-term Nvidia price swings, the source article also said MEXC offers NVDA USDT-M stock futures. It added that stock futures are derivatives and do not represent ownership of Nvidia shares.
Questions highlighted in the source FAQ
Did Nvidia beat expectations?
The source article said Nvidia’s fiscal 2027 second-quarter revenue was $96.2 billion, non-GAAP diluted EPS was $2.22, and data center revenue reached $89 billion. It described the results as above market expectations, while noting that investor attention quickly shifted to Rubin, future AI demand and margins.
What is Nvidia’s fiscal third-quarter revenue guidance?
Nvidia expects about $108 billion in fiscal 2027 third-quarter revenue, plus or minus 2%. The company said that forecast excludes China data center compute revenue.
Why does Vera Rubin matter for Nvidia’s stock?
The source article described Vera Rubin as Nvidia’s next-generation AI computing platform, now in full production. Its deployment progress will help the market judge whether the rapid AI infrastructure expansion seen during the Blackwell cycle can continue into the next generation of Nvidia hardware.
Why is gross margin declining?
Nvidia expects roughly 74% gross margin in the third quarter, down from 75% in the second quarter. The source article said higher memory and other component costs, along with the growing complexity of full AI systems, are becoming important margin drivers.
What should investors watch after earnings?
The source article listed Rubin production and deployment speed, hyperscaler and AI lab capital spending, gross margin trends, memory and component costs, and any change in Nvidia’s ability to sell data center products into China.