NVIDIA (NVDA): 11/3/26

NVIDIA Corporation (NVDA)

11/3/26

Nvidia supplies the technology behind much of today’s AI computing infrastructure. Its offerings extend beyond graphics processing units, or GPUs, to include processors, complete computing systems, networking, and software. Together, these products support both AI training—the development of a model—and inference, when that model generates responses or performs tasks.

The company also serves gaming, professional visualization, scientific research, robotics, and autonomous driving. Nevertheless, its investment outlook increasingly depends on demand for large-scale AI infrastructure.

Fiscal 2026 revenue totaled $215.9 billion, compared with $130.5 billion a year earlier, an increase of approximately 65%. The fiscal year ended January 25, 2026. Compute & Networking accounted for approximately 90% of revenue, with Graphics contributing the remaining 10%.

Beginning in fiscal 2027, Nvidia also presented its business through two market platforms: Data Center and Edge Computing. These categories supplement its reporting and do not replace its accounting segments.

Data Center represents approximately 91% of revenue. It includes computing, networking, software, and services used for AI, analytics, and scientific workloads. Within this platform, Hyperscale serves major cloud and consumer-internet companies, including Amazon, Google, and Microsoft. AI Clouds, Industrial and Enterprise, or ACIE, covers specialized AI cloud providers (neoclouds such as CoreWeave), enterprises, industrial customers, and national AI initiatives. Hyperscale contributed approximately 54% of Data Center revenue, while ACIE represented 46%.

Edge Computing accounts for approximately 9% of revenue and encompasses applications such as gaming, workstations, automotive systems, robotics, and telecommunications. These technologies bring computing closer to where information is generated or used.

Investment thesis

My positive outlook centers on Nvidia’s ability to capture a larger share of spending on complete AI systems. Selling processors alongside networking, software, and supporting infrastructure expands its opportunity beyond individual GPU purchases. However, business growth alone does not ensure attractive stock returns; valuation and the sustainability of customer investment remain essential considerations.

An integrated platform strengthens its competitive position

Nvidia’s advantage combines hardware performance with an established software ecosystem. CUDA allows developers to build and optimize applications for its GPUs. Moving those applications to another platform can require substantial engineering work, testing, and expense, giving customers a reason to remain within Nvidia’s ecosystem.

This creates meaningful switching costs, although it does not eliminate competition. Nvidia must continue improving performance, efficiency, and overall system costs to maintain its position.

Networking adds another layer to the offering. Large AI systems require thousands of processors to communicate efficiently, making interconnects and networking equipment important components of system performance. Nvidia’s ability to supply these alongside its processors helps it compete for a greater portion of each infrastructure project.

Software provides additional revenue opportunities. Many developer tools are freely available, while commercial offerings such as Nvidia AI Enterprise generally involve subscriptions, sometimes bundled with hardware.

Data Center growth extends across customer groups

Second-quarter fiscal 2027 Data Center revenue reached $89 billion, increasing 117% year over year and 18% from the previous quarter. Hyperscale revenue totaled $48.7 billion, up 102%, while ACIE revenue reached $40.3 billion, up 138%.

These results suggest that demand extends beyond the largest cloud companies to specialized AI providers, enterprises, and sovereign customers. Networking revenue also reached a quarterly record, rising 18% sequentially, while Spectrum-X Ethernet revenue increased 2.6 times year over year.

Management expects neocloud partners to finish fiscal 2027 with eight gigawatts of installed capacity, compared with approximately three gigawatts at the end of 2025. Nvidia’s third-quarter fiscal 2027 revenue guidance is $108 billion, plus or minus 2%, excluding any assumed revenue from China Data Center compute sales.

Management also projects approximately 70% revenue growth in fiscal 2028 while expecting supply constraints to persist. These projections support the growth thesis, but their realization depends on customer financing, infrastructure availability, and sustained demand.

Usage-based arrangements could broaden revenue sources

Certain cloud agreements could allow Nvidia to participate in revenue generated after its equipment is sold. Under these arrangements, a provider purchases Nvidia infrastructure and rents computing capacity to customers, with Nvidia potentially receiving a portion of the resulting revenue.

Some structures also involve Nvidia committing to purchase a portion of capacity, establishing a minimum revenue floor that can help the provider obtain financing. Nvidia may then share in rental revenue above that threshold.

Such agreements could create recurring income linked to AI usage. However, the purchase commitments also introduce obligations and financial exposure. Their economic value depends on utilization, pricing, contract terms, and the 

costs Nvidia assumes. Management’s discussion of a potential multibillion-dollar opportunity should therefore be treated as a future possibility rather than assured revenue.

New platforms could improve customer economics

Nvidia’s product development schedule aims to deliver more useful AI output within customers’ power and cost constraints. Vera Rubin began production shipments in August 2026, supported by orders from major hyperscalers, AI clouds, and system manufacturers.

According to management, Rubin can deliver 30 times greater throughput per megawatt and 35 times lower cost per token than Grace Blackwell Ultra. Management also estimates that the platform increases Nvidia’s revenue opportunity per gigawatt from approximately $25 billion with Blackwell to $40 billion.

These performance claims depend on the workloads and configurations used for comparison. Their investment significance is that better computing economics could encourage customers to adopt newer systems.

The broader platform combines the Vera CPU, Rubin GPU, networking, and other processing technologies. Nvidia also expects CPU revenue to more than double in fiscal 2028, providing another potential source of growth.

Agents, robotics, and national AI projects expand potential demand

AI agents could increase computing requirements by completing multistep tasks rather than simply responding to individual questions. Management estimates that agentic workloads can require 15–100 times more computing per user than conventional AI queries, depending on the application.

Physical AI offers a further opportunity through robotics, autonomous vehicles, and industrial automation. Nvidia’s Cosmos, Omniverse, Isaac, and Alpamayo technologies support development in these areas.

Sovereign AI initiatives could also diversify demand. Nvidia reported that its sovereign business grew 35% sequentially and more than tripled year over year in the July quarter. Management expects ACIE customers to account for approximately half of Data Center revenue going forward.

Risks to the outlook

Infrastructure spending may outrun profitable demand

The largest risk is that computing capacity expands faster than customers’ willingness to pay for AI services. Strong adoption does not necessarily produce adequate returns on infrastructure investment.

If supply exceeds profitable demand, cloud providers could face lower utilization and pricing pressure. Depreciation, leases, staffing, and financing commitments would continue even as revenue disappoints. Customers could respond by delaying projects or reducing equipment purchases, weakening Nvidia’s growth.

Financing arrangements create additional exposure

Nvidia’s investments, capacity commitments, and revenue-sharing agreements may help customers fund infrastructure projects. They also connect 

Nvidia more closely to those customers’ financial performance.

Hardware sales and subsequent usage revenue should therefore be evaluated alongside the commitments required to support them. Weak customer economics could limit the benefits of these arrangements.

China restrictions and competition could weaken its position

Restrictions on sales to China constrain Nvidia’s access to a major market. They may also encourage customers and developers to adopt domestic alternatives, helping competing ecosystems mature.

Over time, those competitors could challenge Nvidia beyond China. Its existing advantages require continued investment and execution across processors, software, networking, and complete systems.

Rising expenses increase sensitivity to slower growth

Second-quarter fiscal 2027 non-GAAP operating expenses rose 54% year over year to $8.2 billion. Management expects approximately $9 billion in the third quarter and guides to a non-GAAP gross margin of 74%, plus or minus 50 basis points, compared with 75% in the second quarter.

These investments support product development and future growth. However, if revenue falls short of expectations, a larger expense base could place greater pressure on profitability.

Valuation can limit stock returns

Nvidia can remain a leading AI supplier while its shares deliver disappointing returns. If the stock price already assumes exceptional growth, even strong results may be insufficient to exceed investor expectations. 

My positive outlook therefore depends on both continued business execution and a valuation that reasonably reflects the risks.

Financial Results and Outlook

NVDA recorded revenue of $96.22 billion in the second quarter of fiscal 2027. Revenue for the quarter exceeded guidance of $91 billion and was up 17.91% from Q1 and up 106% from the same quarter in the previous year. The increase came mostly from data center revenue of $89 billion, up 18.35% from Q1 and up 116.5% from a year ago. 

In Q2FY27, GAAP diluted EPS was $2.46, up 2.92% from the previous quarter and up 127.78% from a year ago. Free cash flow was $21.3 billion in the second quarter of fiscal 2027, up from $13.5 billion a year earlier. Nvidia expects Q3 FY2027 revenue of $108 billion, ±2%, nearly double the revenue reported in the same quarter a year earlier.

Valuation

NVDA stock has returned 10.4% year to date, underperforming the S&P 500, which is up 11.56%. The company’s forward GAAP P/E ratio of 20.47x is below the IT sector median of 28.79x. This indicates that investors are paying 28.92% less for each dollar of annual earnings of NVDA stock compared to the IT sector median.

NVDA’s Forward Non-GAAP PEG (Price/Earnings to Growth) ratio of 0.44 suggests that the stock is trading at a discount to its earnings per share growth rate. Additionally, it indicates that investors buying NVDA are paying less per unit of earnings growth compared to the IT sector median PEG ratio of 1.13.

NVDA’s forward Price/Book ratio of 16.50 is 259.55% higher than the IT sector median of 4.59, suggesting that investors have high expectations for NVDA stock because they think it will earn and grow more in the future. 

Lastly, NVDA’s forward price-to-sales ratio of 12.00 is 265.42% higher than the IT sector median of 3.28, implying that investors are paying $12 for every one dollar of Nvidia’s sales.