NVIDIA Corporation (NVDA) Report Interpretation

Overview

J.P. Morgan summarizes investor-meeting takeaways from NVIDIA's investor-relations team and reiterates a constructive view. The report centers on broad AI-compute demand, growing inference exposure, supply constraints, customer diversification, and NVIDIA's integrated platform advantage.

Core views

NVIDIA management described its FY28 70% year-on-year growth framework as supported by broad-based improvement in demand and visibility rather than one factor. Demand spans hyperscalers, neoclouds, AI labs, sovereign AI, and enterprise/on-premise deployments. Management said the out-year outlook also addressed a meaningful gap between Street estimates and its internal view that could otherwise complicate partners' supply-chain planning. It characterized the outlook as supply-constrained rather than demand-constrained and said the business could more than double year on year absent supply limitations. Within the data-center business, NVIDIA said inference is now larger than training and should continue increasing as a share of revenue. It could not precisely quantify the mix because its platforms are fungible: customers can use Grace Blackwell products for training and later redirect the same assets to inference. The mix was approximately 50/50 around 18 months ago. Management also sees a wider demand ecosystem: OpenAI and Anthropic together account for roughly 20% of business on an end-consumption basis today and could approach 25% in FY28, while neoclouds represent more than approximately 50% of ACIE. This supports the view that growth is not dependent solely on the largest hyperscalers. The main supply-side constraints are advanced wafers and memory, both important components of NVIDIA's bill of materials. NVIDIA is engaging with TSMC and memory suppliers Micron, SK Hynix, and Samsung to improve availability. The report also notes that both open-source and closed-source models are likely needed for further computing advances. As model builders monetize products and improve their economics, including through lower generation-on-generation cost per token enabled by NVIDIA compute, management expects demand to flow through to chip providers. NVIDIA outlined financing arrangements intended to support future AI-infrastructure demand, 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. Under revenue-share arrangements, NVIDIA backstops a base price and shares upside when compute is leased above market rates, creating potential recurring revenue in addition to hardware sales. Management argued that these agreements are measured and capped, supported by underlying demand, ecosystem returns, and the credit profiles of ultimate compute purchasers. J.P. Morgan's investment thesis emphasizes 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 several incremental revenue opportunities and is supplemented by a structural shift of traditional enterprise workloads to accelerated computing. The $320 price target assumes approximately 20x CY26 EPS of $15.87. Key challenges are alternative compute platforms capturing AI-compute TAM share—GPU and ASIC/XPU share is expected to trend toward parity in coming years—and memory-cost inflation and supply constraints that may pressure platform specifications, revenue mix, and gross-margin durability.

Analysis framework

The report synthesizes comments from a virtual investor meeting with NVIDIA's investor-relations executive. It evaluates the demand outlook by customer type, examines training-versus-inference mix and supply bottlenecks, discusses financing structures supporting infrastructure demand, and links these observations to NVIDIA's platform advantage and a price-target valuation based on a multiple of CY26 EPS.

Methodology notes

• Valuation methodsP/E and PEG Valuation

  Price-to-earnings multiple valuation

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

• Industry AnalysisSupply-demand framework

  AI-compute supply-demand analysis

  The report assesses demand across customer groups against advanced-wafer and memory constraints to explain why the FY28 outlook is supply-constrained rather than demand-constrained.

Asset mapping & comparison

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

• NVIDIA Corporation (NVDA)

  Primary covered company; the report views broad AI-compute demand, platform integration, and increasing inference exposure as supportive.

  Strengths

  Vertically integrated chips, rack systems, and software stack; strong multi-year demand visibility; broadening customer ecosystem.

  Weaknesses

  Supply is constrained by advanced wafers and memory, limiting the ability to meet demand.

  Comparison

  Alternative AI ASICs/XPUs, merchant competitors including AMD, and specialized accelerators may gain TAM share; GPU and ASIC/XPU shares are expected to move toward parity over time.

  Risks

  Memory cost inflation and supply constraints may pressure specifications, mix, and gross-margin durability.

Key data

• FY28 growth framework70% Y/YManagement said it is supported by broad-based demand and greater visibility.
• Potential growth absent supply constraintsMore than double Y/YManagement's indication of demand potential.
• Training/inference mix around 18 months ago~50/50Inference is now larger and expected to continue growing as a share of business.
• OpenAI and Anthropic end-consumption share~20% today; potentially ~25% in FY28Direct customer mix may differ because NVIDIA sells through hyperscalers and neoclouds.
• Neocloud share of ACIEMore than ~50%Management's indication that the customer base is broadening.
• Private-capital financing platform$500BPlatform cited as supporting forward AI-infrastructure demand.
• CY26 EPS used for valuation$15.87J.P. Morgan applies an approximately 20x multiple.

Impact & implications

The report says demand is broadening across AI labs, neoclouds, sovereign AI, enterprises, and hyperscalers, while supply availability limits near-term fulfillment. It views NVIDIA's integrated hardware-and-software platform and increasing inference exposure as supportive, while recognizing that alternative accelerators and memory economics could challenge market share and profitability.

Risks

• Alternative compute platforms, including AI ASICs/XPUs, AMD, and specialized accelerators, could capture AI-compute TAM share.
• Memory cost inflation and supply constraints could pressure platform specifications and create uncertainty over revenue mix and gross-margin durability.