Nvidia expects its revenue to grow by approximately 70% in fiscal 2028, a forecast that underscores the extraordinary strength of demand for artificial intelligence infrastructure. The company’s chief financial officer, Colette Kress, said the outlook is “supply-constrained,” meaning Nvidia believes it could grow even faster if it had enough components and manufacturing capacity to satisfy customer demand.
The projection is significantly above Wall Street expectations of roughly 44–45% growth for fiscal 2028. Nvidia’s forecast came after the company reported $96.2 billion in second-quarter fiscal 2027 revenue, up 106% year over year, and guided for $108 billion in third-quarter revenue. The company said demand is accelerating across hyperscalers, AI labs, startups, enterprises and sovereign customers, while shortages — particularly in memory — are limiting how quickly it can expand supply.
Nvidia Expects 70% Revenue Growth In Fiscal 2028
The fiscal 2028 outlook is unusual for Nvidia because the company typically provides quarterly guidance rather than a detailed year-ahead revenue forecast.
Kress said Nvidia’s preliminary expectation is for revenue to increase by approximately 70% year over year in fiscal 2028. She also emphasized that this is not a demand-limited forecast; rather, supply availability is expected to remain a constraint.
Nvidia’s Growth Outlook
| Metric | Figure |
|---|---|
| Q2 FY2027 revenue | $96.2 Bn |
| Q2 FY2027 YoY growth | 106% |
| Q2 Data Center revenue | $89.0 Bn |
| Q2 Data Center YoY growth | 117% |
| Q3 FY2027 revenue guidance | $108 Bn ±2% |
| Fiscal 2028 revenue growth outlook | ~70% |
| Wall Street FY2028 growth expectation | ~44–45% |
| FY2028 outlook | Supply constrained |
The gap between Nvidia’s own outlook and the market consensus is significant and suggests that management sees AI infrastructure demand remaining unusually strong well beyond the current fiscal year.
What Does “Supply-Constrained” Mean?
Nvidia’s statement is important because it changes the interpretation of the 70% forecast.
The company is not saying that demand is expected to grow by only 70%.
Instead, management indicated that customer forecasts point to growth potentially doubling next year, but Nvidia expects to deliver approximately 70% growth because supply will remain a bottleneck.
Customer AI Demand
↓
Potential Growth: ~100%
↓
Nvidia Supply Capacity
↓
Memory + Manufacturing Constraints
↓
Expected Revenue Growth
↓
~70%Kress said Nvidia expects supply to remain a bottleneck at least through the end of fiscal 2028, even as the company works to close the supply-demand gap.
Demand Is Growing Faster Than Nvidia Can Supply
The company’s comments indicate that the bottleneck is not a lack of customer demand.
Nvidia said its compute capacity is fully utilized across every cloud it serves, while customers’ forecasts point to growth potentially doubling the following year.
This is an unusual situation for a semiconductor company: instead of trying to stimulate demand, Nvidia is trying to increase manufacturing capacity quickly enough to fulfill orders already being generated by customers.
Demand-Supply Dynamic
| Factor | Current Situation |
|---|---|
| AI compute demand | Accelerating |
| Customer forecasts | Point toward potentially 2x growth |
| Nvidia supply | Constrained |
| Cloud capacity | Fully utilized |
| Memory availability | Tight |
| Manufacturing capacity | Major bottleneck |
| FY2028 revenue outlook | ~70% growth |
The situation suggests that Nvidia’s near-term ceiling may be determined more by production capacity than by customer appetite.
AI Demand Is Expanding Beyond Hyperscalers
Another reason Nvidia is confident about future growth is that demand is becoming more diversified.
Historically, Nvidia’s AI growth was heavily associated with major cloud companies and hyperscalers.
Now, the company says demand is coming from a much wider group of customers.
Hyperscalers
+
AI Labs
+
AI Startups
+
NeoClouds
+
Enterprises
+
Sovereign Customers
+
Industrial / Physical AI
↓
Broader AI Compute DemandKress said the global AI infrastructure buildout is being supported by hyperscalers, AI labs, AI natives, enterprises and sovereign customers.
This diversification could make Nvidia’s growth less dependent on a small number of customers.
Data Center Revenue Reaches $89 Billion
Nvidia’s second-quarter results provide the financial foundation for the 70% fiscal 2028 outlook.
Data Center revenue reached $89 billion, up 117% year over year and 18% sequentially.
The segment now represents more than 92% of Nvidia’s total quarterly revenue.
Nvidia Data Center Growth
| Metric | Q2 FY2027 |
|---|---|
| Data Center revenue | $89 Bn |
| YoY growth | 117% |
| Sequential growth | 18% |
| Share of Nvidia revenue | ~92.5% |
This demonstrates how deeply Nvidia’s financial performance is now tied to AI infrastructure spending.
Nvidia’s Q3 Revenue Could Cross $100 Billion
Nvidia expects third-quarter revenue of $108 billion, plus or minus 2%.
That represents a potential quarterly revenue range of approximately $105.84 billion to $110.16 billion.
If Nvidia reaches the midpoint, it would be the company’s first quarter above the $100 billion revenue threshold.
Q2 FY2026
$46.7 Bn
↓
Q2 FY2027
$96.2 Bn
↓
Q3 FY2027 Guidance
$108.0 Bn
↓
FY2028 Outlook
~70% Revenue GrowthThe trajectory highlights how quickly Nvidia’s financial scale has expanded alongside AI investment.
Vera Rubin Could Become A Major Growth Driver
Nvidia’s next-generation Vera Rubin platform is expected to play an important role in fiscal 2028 growth.
The company said Vera Rubin has started shipping to customers and expects it to represent approximately 20% of Data Center revenue in the current quarter.
Nvidia expects hyperscale growth to reaccelerate in Q4 and into fiscal 2028 as Vera Rubin supply increases.
Nvidia Platform Transition
Blackwell
↓
Vera Rubin
↓
Higher AI Compute Density
↓
More Training + Inference
↓
Larger Customer Deployments
↓
Fiscal 2028 Revenue GrowthThe faster Nvidia can ramp Rubin systems, the more of the existing demand backlog it can potentially convert into revenue.
Memory Shortages Are The Biggest Constraint
The most significant supply problem Nvidia highlighted is memory.
Kress said Nvidia is experiencing “extreme pricing conditions in memory”, with price increases exceeding previous expectations and expected to move higher into the next year.
AI accelerators require high-performance memory, including high-bandwidth memory (HBM).
As AI infrastructure expands, demand for these components has increased rapidly.
Memory Supply Chain
| Stage | Impact |
|---|---|
| AI infrastructure demand rises | More accelerators required |
| Accelerator production increases | More HBM required |
| HBM demand rises | Memory suppliers face capacity pressure |
| Memory prices increase | Nvidia’s component costs rise |
| Supply remains tight | Nvidia’s growth becomes constrained |
Nvidia said it has long-standing relationships with the three major memory suppliers and is working with them to increase capacity.
Memory Costs Will Pressure Nvidia’s Margins
The supply problem has a financial consequence.
Nvidia expects gross margins to decline from 75% in Q2 to approximately 74% in Q3, then bottom at around 71–72% in Q4 before settling at approximately 72–73% in fiscal 2028.
Nvidia Gross Margin Outlook
| Period | Gross Margin |
|---|---|
| Q2 FY2027 | 75% |
| Q3 FY2027 | ~74% |
| Q4 FY2027 | ~71–72% |
| FY2028 | ~72–73% |
Nvidia expects executed price increases beginning in the first quarter of fiscal 2028 to help offset higher component costs.
Nvidia May Raise Prices To Offset Component Inflation
Nvidia’s ability to maintain strong margins despite rising memory costs will depend partly on pricing power.
The company expects price increases to take effect in fiscal 2028.
This means customers could face higher prices for Nvidia’s AI systems even as Nvidia expands production.
Memory Prices Rise
↓
Higher Nvidia Component Costs
↓
Gross Margin Pressure
↓
Nvidia Price Increases
↓
Partial Cost Recovery
↓
Margins StabilizeThe pricing response also indicates that demand remains strong enough for Nvidia to believe customers will continue purchasing despite higher system costs.
AI Infrastructure Spending Shows No Sign Of Slowing
Nvidia’s long-term forecast comes as hyperscalers continue spending heavily on AI infrastructure.
According to Nvidia, the cloud industry’s backlog is now more than $2 trillion, while capital expenditures by the five largest hyperscalers are expected to reach nearly $800 billion in 2026 and $1.3 trillion in 2027.
Hyperscaler Investment
| Metric | Outlook |
|---|---|
| Cloud industry backlog | >$2 Trillion |
| Top-five hyperscaler CapEx, 2026 | ~$800 Bn |
| Top-five hyperscaler CapEx, 2027 | ~$1.3 Tn |
| Nvidia FY2028 revenue growth | ~70% |
These figures provide context for Nvidia’s confidence that AI infrastructure demand can remain elevated for several years.
AWS Adds 2 Million Nvidia GPUs
Nvidia also announced an expanded partnership with Amazon Web Services (AWS).
AWS plans to deploy an additional 2 million Nvidia GPUs starting in the current quarter and continuing through the second quarter of fiscal 2029. The deployment will also include Vera CPUs, with some integrated with Rubin systems.
AWS-Nvidia Expansion
| Item | Details |
|---|---|
| Additional GPUs | 2 million |
| Deployment start | Current quarter |
| Deployment period | Through Q2 FY2029 |
| CPU component | Nvidia Vera |
| Rubin integration | Included in some systems |
The agreement provides a visible source of future demand for Nvidia’s computing platforms.
AI Labs Could Become A Larger Customer Base
Nvidia also expects AI labs to contribute approximately one-quarter of its overall business next year, according to Reuters.
This is important because frontier AI laboratories are among the most compute-intensive customers in the industry.
Companies developing increasingly capable models require enormous quantities of accelerators for both training and inference.
AI Lab Demand
Frontier AI Models
↓
More Training
+
More Inference
↓
More GPU Clusters
↓
Higher Nvidia Demand
↓
Potentially ~25%
of Nvidia BusinessThe growing contribution from AI labs could become a major driver of Nvidia’s fiscal 2028 revenue.
NeoClouds Are Expanding Rapidly
Nvidia is also benefiting from the growth of specialized AI cloud providers, sometimes called NeoClouds.
Companies such as CoreWeave and Nebius are adding large amounts of Nvidia GPU capacity to serve AI startups, enterprises and sovereign customers.
Nvidia said these NeoCloud partners are expected to exit 2026 with approximately 8 gigawatts of installed capacity, compared with around 3 gigawatts at the end of 2025.
NeoCloud Capacity
| Period | Nvidia GPU Capacity |
|---|---|
| End-2025 | ~3 GW |
| Expected end-2026 | ~8 GW |
| Increase | ~167% |
The rapid expansion illustrates the broadening of AI compute demand beyond traditional hyperscalers.
Supply Constraints Could Actually Strengthen Nvidia’s Pricing Power
At first glance, supply constraints appear negative because Nvidia cannot fulfill every potential order.
But scarcity can also strengthen pricing power.
When customers compete for limited supply, Nvidia can potentially increase prices while maintaining high utilization.
This could partly offset higher memory and component costs.
Supply Constraint Effect
Demand > Supply
↓
Limited Availability
↓
High Utilization
↓
Pricing Power
↓
Higher Selling Prices
↓
Cost Inflation Partially OffsetThe key risk would arise if supply constraints eventually ease while customer demand slows at the same time.
Nvidia’s Biggest Risk Is No Longer Demand Alone
The 70% outlook changes the central question surrounding Nvidia.
Previously, investors were increasingly asking whether AI spending would slow.
Now, Nvidia’s management is effectively saying that demand is strong enough to support much faster growth — and the company is constrained by its ability to manufacture enough products.
That shifts the key risks toward:
- Semiconductor supply
- HBM availability
- Manufacturing capacity
- Data-center construction
- Energy availability
- Export restrictions
- Customer concentration
- AI investment returns
China Remains A Major Uncertainty
China is another important variable in Nvidia’s outlook.
The company’s Q3 revenue guidance does not assume Data Center compute revenue from China.
This means the 70% fiscal 2028 outlook is not dependent on a major recovery in Nvidia’s China data-center business.
However, restrictions on advanced AI chips remain an important geopolitical risk.
Nvidia Is Expanding Beyond GPUs
Nvidia’s growth strategy is also becoming broader than its traditional GPU business.
The company is investing in:
- AI CPUs
- Networking
- Inference accelerators
- AI software
- AI agents
- Robotics
- Physical AI
- Sovereign AI infrastructure
The Vera Rubin platform is central to this strategy.
The company is effectively trying to become the full-stack infrastructure provider for AI rather than simply the world’s largest GPU supplier.
The 70% Forecast Is A Statement About AI’s Duration
Perhaps the biggest significance of Nvidia’s fiscal 2028 forecast is that it pushes the expected duration of the AI infrastructure boom further into the future.
A 70% growth outlook means Nvidia does not expect AI infrastructure investment to normalize rapidly after the current wave of data-center construction.
Instead, the company sees multiple stages of AI adoption developing simultaneously.
AI Training
↓
AI Inference
↓
AI Agents
↓
Enterprise AI
↓
Sovereign AI
↓
Physical AI
↓
Robotics
↓
Autonomous SystemsEach new workload creates additional demand for computing capacity.
What It Means For Investors
For investors, Nvidia’s forecast is both bullish and cautionary.
The bullish argument is straightforward: if Nvidia can grow revenue by roughly 70% in fiscal 2028, the company’s earnings potential remains far above what the market had expected.
The caution is that the forecast depends on an exceptionally strong AI spending environment continuing for another year and on Nvidia successfully navigating its supply constraints.
What It Means For The AI Industry
For the broader AI industry, Nvidia’s comments indicate that infrastructure investment remains in an expansion phase.
Cloud companies, AI laboratories, enterprises and governments are all competing for compute.
The shortage of advanced memory and AI accelerators could therefore remain an industry-wide issue through at least 2028.
AI Infrastructure Outlook
| Area | Outlook |
|---|---|
| AI compute demand | Very strong |
| Nvidia revenue growth FY2028 | ~70% |
| Supply | Constrained |
| Memory | Tight and expensive |
| AI data centers | Rapid expansion |
| Hyperscaler CapEx | Rising |
| AI labs | Increasing contribution |
| Rubin platform | Major growth driver |
The Bigger Picture
Nvidia’s preliminary forecast for approximately 70% revenue growth in fiscal 2028 is one of the strongest signals yet that the AI infrastructure boom remains far from mature. More importantly, Nvidia says the figure is a supply-constrained outlook. Customer forecasts point to potentially doubling growth, but shortages in memory and other components mean Nvidia expects to capture only about 70% revenue growth. The company expects supply to remain a bottleneck at least through the end of fiscal 2028.
The forecast is especially significant because it is far above the roughly 44–45% growth analysts had expected. Nvidia is also entering this period with expanding demand from hyperscalers, AI labs, NeoClouds, enterprises and sovereign customers. AWS plans to deploy an additional 2 million Nvidia GPUs, while NeoCloud capacity is expected to grow from about 3 gigawatts at the end of 2025 to 8 gigawatts by the end of 2026.
Looking Ahead
Nvidia’s biggest challenge may now be converting unprecedented demand into physical supply. The company is working with all three major memory suppliers to expand capacity, while its Vera Rubin platform is ramping and expected to account for about 20% of Data Center revenue in the third quarter. Nvidia also expects price increases in fiscal 2028 to help offset higher memory costs, with gross margins projected to settle around 72–73% after bottoming at 71–72% in Q4.
The broader implication is that Nvidia believes AI infrastructure spending can remain a high-growth market through at least 2028. If supply expands fast enough, the company’s revenue could potentially exceed the current 70% outlook; if demand remains strong but component shortages persist, the constraint could continue to cap sales. For investors, the crucial variables will therefore be the pace of AI demand, Nvidia’s ability to increase supply, memory costs, pricing power and whether the enormous capital being deployed into AI infrastructure continues to generate sufficient economic returns
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
