NVIDIA Corporation (NVDA) Report Interpretation

Overview

This investor-meeting note summarizes NVIDIA management's demand, supply, customer-mix and financing commentary. J.P. Morgan maintains an Overweight view, arguing that the company's FY28 outlook is underpinned by broad AI-compute demand and greater forward visibility, while supply availability remains the principal operating constraint.

Core views

Management said its FY28 framework of 70% year-on-year growth is supported by strength across hyperscalers, neoclouds, AI labs, sovereign AI and enterprise/on-premise demand rather than any single driver. NVIDIA provided an out-year outlook partly because its internal view was meaningfully above Street estimates, which it believed could otherwise complicate supply-chain planning for partners. The company characterized the outlook as supply-constrained rather than demand-constrained: absent supply constraints, management said the business could more than double year on year. Within Data Center, management said it is difficult to measure training and inference revenue precisely because NVIDIA platforms are fungible: customers can use Grace Blackwell systems for training and later shift the same assets to inference. Still, the mix was roughly 50/50 around 18 months ago, and management now believes inference is larger and should continue to become a greater portion of the business. This shift is supported by demand from model builders whose economics appear to be improving as NVIDIA's generation-to-generation cost-per-token declines help raise gross profit per token. Customer demand is broadening beyond the largest hyperscalers. OpenAI and Anthropic together account for about 20% of NVIDIA's business on an end-consumption basis today and could move toward about 25% in FY28, although this is not necessarily visible in NVIDIA's direct customer mix because compute is sold through hyperscalers and neoclouds. Neoclouds account for more than about 50% of ACIE, according to management, supporting J.P. Morgan's view that growth increasingly reflects a wider ecosystem building or leasing compute capacity. The principal supply-side constraints are advanced wafers and memory, both important components of NVIDIA's bill of materials. Management said it is working closely with TSMC and memory suppliers Micron, SK Hynix and Samsung to improve availability. The same bottlenecks create a key financial risk: memory-cost inflation and supply limitations have pressured platform specifications, could affect product mix and profitability, and raise uncertainty about the durability of gross margins as NVIDIA scales back memory capacity to meet compute-volume commitments. Management also addressed the open-source versus closed-source model debate, arguing that continued AI progress will require both approaches. NVIDIA itself uses closed models including OpenAI and Claude, while employing a mix of open and closed models for mission-critical chip-design work. The relevant demand mechanism, in management's view, is model-builder monetization and improving economics: if those remain intact, compute demand should flow through to providers such as NVIDIA. NVIDIA's financing initiatives—including revenue-share agreements with selected neoclouds, the PORTS-Pike data-center campus initiative, and a $500 billion private-capital financing platform involving Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, KKR and others—are intended to support forward AI-infrastructure demand. Under revenue-share agreements, NVIDIA helps backstop a base price and shares in upside when compute is leased above market assumptions, creating recurring-revenue optionality alongside hardware sales. Management rejected circular-financing concerns, describing the arrangements as measured and capped, supported by demand, ecosystem returns and the credit quality of ultimate compute purchasers. J.P. Morgan's investment thesis rests on NVIDIA's vertically integrated platform across chips, rack systems and software, which it views as difficult for AI-compute peers to replicate. It sees multi-year demand visibility as a floor that excludes some incremental revenue streams and is supplemented by a structural shift of traditional enterprise workloads toward accelerated computing. The firm values NVIDIA at about 20x CY26 EPS of $15.87 to derive its $320 price target. It also flags a competitive headwind: alternative AI-compute platforms, especially AI ASICs and XPUs as well as merchant competitors such as AMD and specialized accelerators, are expected to move toward parity with GPUs in AI-compute total addressable market share over coming years.

Analysis framework

The note draws on a virtual investor discussion with NVIDIA's Vice President of Investor Relations & Strategic Finance. It evaluates the FY28 outlook by tracing demand across customer groups, testing whether supply rather than demand limits growth, assessing the training-to-inference mix, identifying component bottlenecks, and reviewing customer financing. J.P. Morgan then connects these operating observations to its integrated-platform thesis and a CY26 EPS multiple-based price target.

Methodology notes

• Valuation methodsP/E and PEG Valuation

  Price-to-earnings valuation

  J.P. Morgan applies an approximately 20x multiple, reflecting long-term growth expectations, to CY26 EPS of $15.87 to derive its $320 price target.

• Industry AnalysisSupply-demand framework

  AI-compute demand versus component supply constraints

  The report distinguishes broad demand across customer groups from limited advanced-wafer and memory availability to explain why the FY28 outlook is supply-constrained rather than demand-constrained.

• Competition & strategyEconomic Moat and Competitive Advantage

  Vertically integrated AI-compute platform

  The investment thesis treats NVIDIA's combination of chips, rack systems and software as difficult for peers to replicate, while recognizing alternative AI-compute platforms as a competitive risk.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• NVIDIA Corporation (NVDA)

  Primary covered company; positioned as a beneficiary of broad AI-compute demand and accelerating-compute adoption.

  Strengths

  A vertically integrated platform spanning chips, rack systems and software; strong multi-year demand visibility; increasing inference exposure.

  Weaknesses

  Platform specifications and profitability face pressure from memory cost inflation and supply constraints.

  Comparison

  J.P. Morgan sees the platform as difficult to replicate versus AI-compute peers, but expects GPU and ASIC/XPU AI-compute TAM shares to trend toward parity over coming years.

  Risks

  Alternative compute platforms, memory inflation, advanced-wafer and memory shortages, and uncertainty around long-term gross-margin durability.

Key data

• FY28 growth framework70% Y/YManagement said it is comfortable with this framework; absent supply constraints, the business could more than double Y/Y.
• Training/inference mix~50/50 around 18 months agoManagement believes inference is now larger and will continue to grow as a share of the business.
• Frontier model-builder end-consumption share~20% today; potentially ~25% in FY28Represents OpenAI and Anthropic together on an end-consumption basis.
• Neocloud share of ACIEMore than ~50%Management cited this as evidence of demand broadening beyond the largest hyperscalers.
• Private-capital financing platform$500BPlatform includes Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, KKR and others.
• Price-target valuation~20x CY26 EPS of $15.87J.P. Morgan's basis for the $320 price target.

Impact & implications

J.P. Morgan argues that NVIDIA's growth is supported by a wider and more visible AI-compute demand base, with inference, neoclouds and enterprise workloads adding to hyperscaler demand. The report's central constraint is supply, particularly memory and advanced wafers; competitive alternative accelerators and potential gross-margin pressure remain important offsets to the constructive thesis.

Risks

• AI ASICs, XPUs, AMD and specialized accelerators could capture AI-compute total addressable market share, with GPU and ASIC/XPU shares expected to trend toward parity in coming years.
• Memory cost inflation and supply constraints may pressure platform specifications, product mix and gross-margin durability.
• Advanced-wafer and memory availability may limit NVIDIA's ability to meet demand.

What to watch

• Whether advanced-wafer and memory supply availability improves sufficiently to meet FY28 demand.
• The pace at which inference expands relative to training in NVIDIA's Data Center business.
• Demand growth from neoclouds, frontier model builders, sovereign AI and enterprise/on-premise customers.
• Model-builder monetization and cost-per-token improvements that support downstream compute demand.
• Execution and demand support from NVIDIA's revenue-share and private-capital financing initiatives.