Dell Technologies Inc. (DELL): Dell's Agentic AI infrastructure opportunity looks more durable, but Morgan Stanley retains Equal-weight

Overview

Morgan Stanley summarizes a one-on-one meeting with Dell COO Jeff Clarke and reassesses whether Agentic AI makes the current infrastructure cycle structurally different from prior hardware boom-and-bust periods. The firm sees stronger support for Dell's bull case, but maintains Equal-weight and its $511 target pending greater confidence in the duration of on-premise AI infrastructure demand.

Core views

Morgan Stanley's central conclusion from the meeting is that Agentic AI could enlarge, rather than merely accelerate, the infrastructure market. Dell contrasts this cycle with eight historical hardware cycles, which it characterizes as refresh-driven and therefore prone to reversal once device-per-person penetration normalized. In Dell's view, Agentic AI decouples cognitive output from headcount, creating measurable productivity gains of 10x to 100x and a structural reason for enterprises to spend more on compute and storage. Dell estimates inference-token generation will grow 87x by 2030, alongside 200 incremental gigawatts of demand and a fivefold rise in ZettaFLOPS to 830 by 2030. Because agents require materially more tokens per unit of work than basic chatbots, the report argues that this increases the need to redesign data centers around accelerated compute and higher CPU-to-GPU ratios. The report argues that the benefits should extend beyond cloud providers. Dell expects enterprise inference to be hybrid, with workload placement determined by security and affordability: it uses public cloud applications for content workloads but would retain proprietary source code and telemetry data in its own facilities. Open-weight model adoption could further support on-premise investment as enterprises optimize outputs for cost. This creates what Dell calls a “rising tide lifts all boats” outcome, in which both cloud players and on-premise vendors benefit from a structurally larger compute and storage market. Morgan Stanley also addresses the apparent contradiction between rapid token growth and declining traditional-server units. Newer 17th- and 18th-generation Dell servers can replace up to 13 14th-generation systems, so greater density and efficiency can initially lower unit volumes while lifting average selling prices. Dell expects that, as data centers are rebuilt around accelerated computing and agents raise demand for CPU servers, denser server-unit growth could eventually become exponential. Storage is also positioned as a beneficiary because KV Cache, retained agent information and artifacts increase data-retention requirements. Supply remains the main operational challenge. Dell believes shortages of memory, HDDs and other critical components could persist for more than five years, unlike historical commodity cycles that typically lasted two to four quarters. Dell's logic is that sustained growth in gigawatts and tokens changes component demand from a short cycle into a prolonged scarcity environment. Morgan Stanley highlights Dell's supply-chain management as a source of share gains across servers, storage and PCs, as well as an advantage in serving mission-critical demand and reaching market with leading NVIDIA chips faster than peers. On margins, Dell challenged Morgan Stanley's prior “margin stacking” explanation, under which infrastructure vendors earn a margin on higher memory costs amid inelastic demand. Dell acknowledged that like-for-like server and storage margins are expanding, but argues that scarcity lets it allocate parts to higher-margin products while a structurally growing market reduces the need to price aggressively to win incremental customers. This could raise the infrastructure pricing floor and make margin improvement more structural. Dell is also investing in AI across supply-chain optimization, customer service, coding, sales and R&D. While these investments mean operating expense is unlikely to decline year over year soon, Dell expects operating leverage because revenue grows faster than that spending. The valuation debate remains focused on durability. Morgan Stanley says both bulls and bears expect very strong revenue and earnings growth over the next 18 months; the FY28 buyside EPS benchmark is $40-$50, versus Morgan Stanley's $36.47 and Street consensus of $30.44. Bulls accept the longer-duration infrastructure thesis and apply roughly 15x FY28 EPS, implying a $600-$750 share price; bears expect another boom-bust cycle, believe more AI value will accrue to cloud vendors, and underwrite 8x-10x peak FY28 EPS, implying $320-$500. Morgan Stanley had been closer to the bearish interpretation but is now reconsidering whether token growth can create an elongated on-premise boom. It retains Equal-weight and an unchanged $511 target for now, while underwriting a higher probability that the $756 bull-case valuation is reached within 12 months. The $511 target is based on 14x FY28, or CY27, EPS of $36.47. Morgan Stanley applies 14x to both Dell's AI-server business, in line with AI-infrastructure peers, and its traditional business, about one turn above traditional enterprise-hardware peers because of execution and share gains. Its base case assumes the memory supercycle plays only a minimal role in results over the next two years, followed by a modest ISG correction in FY29. The bull case values FY28 EPS of $42.00 at 18x, assuming demand exceeds supply through CY28/FY29, a three-plus-year spending upcycle, sustained pricing power and EPS growth of more than threefold in two years. The $306 bear case values FY28 EPS of $25.53 at 12x, assuming memory inflation and pull-forward emerge in CY27, PCs decline year over year, server growth rolls over in the second half of CY27, margins come under greater pressure and the P/E multiple contracts.

Analysis framework

Morgan Stanley uses management commentary from its Dell site meeting to test whether AI demand is structural rather than refresh-cycle driven. It links projected inference-token growth to compute, storage, supply and margin effects, then frames the investment debate through base, bull and bear EPS-and-P/E valuation scenarios and compares its estimates with buyside and Street expectations.

Methodology notes

• Industry AnalysisSupply-demand framework

  Infrastructure demand and component-supply scarcity analysis

  The report connects Agentic AI-driven compute and storage demand with prolonged shortages of memory and HDDs to explain Dell's potential share gains and pricing power.

• Valuation methodsP/E and PEG Valuation

  Scenario P/E valuation on FY28 EPS

  Morgan Stanley derives its target and bull, base and bear cases by applying different P/E multiples to FY28 EPS assumptions.

• Event-Driven and Behavioral FinanceEvent-driven analysis

  Management-meeting reassessment

  The report uses Dell management's explanations of AI demand, supply and margins as the event evidence prompting a reassessment of the prior thesis.

Asset mapping & comparison

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

• Dell Technologies Inc. (DELL.N)

  Primary covered company and potential beneficiary of Agentic AI-driven demand for enterprise compute and storage.

  Strengths

  Supply-chain management, share gains in servers, storage and PCs, pricing power, and faster time to market with leading NVIDIA chips.

  Weaknesses

  The durability of AI infrastructure demand and the extent to which on-premise vendors capture AI value remain uncertain.

  Comparison

  Morgan Stanley characterizes Dell as the “best athlete” in its enterprise-infrastructure coverage and values its AI-server business in line with AI-infrastructure peers.

  Risks

  Demand pull-forward, a CY27 server-growth rollover, margin pressure, AI capex slowing, and cloud migration cannibalizing enterprise IT workloads.

Key data

• Price target$511.00Based on 14x FY28 (CY27) EPS of $36.47.
• Current share price$548.92Closing price on Sep 22, 2026.
• Inference-token growth87x by 2030Dell's estimate for Agentic AI inference-token generation.
• FY28 EPS$36.47Morgan Stanley base-case estimate; buyside benchmark is $40-$50 and Street estimate is $30.44.
• Bull/base/bear scenario prices$756 / $511 / $306Bull uses 18x FY28 EPS of $42.00; base uses 14x EPS of $36.47; bear uses 12x EPS of $25.53.
• Revenue growth93.9% / 33.1% / (4.7%)Morgan Stanley estimates for Jan 2027e, Jan 2028e and Jan 2029e, respectively.

Impact & implications

Morgan Stanley sees Dell as a leading enterprise-infrastructure operator because of supply-chain execution, share gains and potential pricing power. However, the report's Equal-weight stance reflects uncertainty over the pace, magnitude and duration of the AI infrastructure buildout, including the possibility that demand has been pulled forward and earnings later roll over.

Risks

• AI infrastructure spending could be pulled forward, causing demand to slow or roll over as the cycle turns in CY27.
• Memory inflation and elevated discounting could pressure margins and trigger concerns that EPS has peaked.
• AI capex could slow, while enterprise IT workloads could migrate to the cloud and reduce on-premise demand.
• Component supply constraints remain Dell's largest operational challenge and may persist for years.

What to watch

• Whether enterprise AI spending sustains infrastructure demand into FY29.
• Server, storage and PC share gains, including Dell's execution in sourcing leading NVIDIA chips.
• The pace and duration of supply scarcity, pricing power and infrastructure-margin expansion.
• Evidence of pull-forward, including PC declines, server-growth deceleration in the second half of CY27 and near-term EPS rolling over.