Dell Technologies Inc. (DELL): Deep Dive Research Report

Prepared: May 24, 2026 | Four concalls used: Q1 FY26 (May 29, 2025), Q2 FY26 (Aug 28, 2025), Q3 FY26 (Nov 25, 2025), Q4 FY26 (Feb 26, 2026)

1. What the Company Does

Dell Technologies is, at its core, a builder and seller of computing infrastructure. It designs, assembles, and services the computers, servers, and storage systems that power almost every facet of the digital economy - from a hospital keeping patient records on a Latitude laptop to a neocloud startup running a trillion-parameter language model on a rack of GPU servers the size of a small car.

The founding story matters here because it explains the company's deepest structural advantage. In 1984, a 19-year-old Michael Dell started building IBM-compatible PCs in his University of Texas dorm room with $1,000 in capital. He called the venture PC's Limited. The insight was simple but radical: cut out the retailer. Sell directly to the customer, build to order, and use just-in-time inventory. By 1987 the company had rebranded to Dell Computer Corporation. By 1996, dell.com was generating $1 million in sales per day. By 2001, Dell was the largest PC vendor in the world.

That direct model - selling to customers, not through middlemen - created two lasting advantages. First, Dell learned to run extremely lean inventory cycles, which meant it could pass component cost deflation to customers faster than competitors and absorb component inflation through rapid pricing adjustments. Second, it built the largest direct enterprise sales force in the industry, with relationships across every Fortune 500 company on the planet.

The second founding moment came in 2015. Dell announced the acquisition of EMC Corporation for $67 billion, the largest technology acquisition in history at that time. EMC was the world leader in enterprise storage - its products included the iconic VMAX storage array and the VMware virtualization platform (which was majority-owned by EMC). The combined entity, renamed Dell Technologies in 2016, became one of the very few technology companies capable of building a complete data center from scratch: servers, storage, networking, virtualization, and services.

Michael Dell stepped off the NYSE in 2013 as part of a leveraged buyout, took the company private, executed the EMC deal, and re-listed on the NYSE in December 2018. The company had been remade.

Today, Dell does one thing with unusual competence: it takes the world's most advanced hardware components - primarily NVIDIA GPUs, Intel and AMD CPUs, DRAM and NAND flash memory, disk drives, and networking silicon - and assembles them into verified, deployable, enterprise-grade systems. The value it adds is not in designing silicon. It is in systems integration: the mechanical engineering of rack-scale cooling systems, the firmware and software stacks that make heterogeneous components work together, the global supply chain that can procure and deliver at scale, and the global services organization that can deploy and keep those systems running. When a customer buys a Dell PowerEdge XE9712 - a rack containing 36 NVIDIA Grace CPUs and 72 Blackwell GPUs interconnected via NVLink - they are not just buying NVIDIA GPUs. They are buying a pre-tested, certify-ready, deployable AI training system with a Dell service team that can turn it on in 24 hours at a customer's data center, anywhere in the world.

"A GB200 NVL72 rack contains 1.2 million parts, requiring a highly complex supply chain orchestration. Our deployment differentiation centers on turning on racks in 24 hours when they're operating."

• Jeff Clarke, Q1 FY26 Concall, May 29, 2025

That sentence captures Dell's current moment. The company is no longer primarily a PC company. It is, by its own trajectory, becoming the world's largest AI infrastructure integrator.

2. Business Segments

Dell operates two reportable segments: Infrastructure Solutions Group (ISG) and Client Solutions Group (CSG). These are genuinely different businesses in terms of customers, economics, competitive dynamics, and strategic role. ISG is the engine of current growth. CSG is the stable foundation.

Infrastructure Solutions Group (ISG)

ISG is Dell's data center and AI infrastructure business. It designs, assembles, and sells servers, storage systems, and networking equipment for enterprises, cloud service providers, sovereign governments, and AI-specialist customers (what Dell calls "neoclouds" - companies like CoreWeave, Voltage Park, and Lambda Labs that rent GPU time to AI researchers and businesses).

Within ISG, there are three product families that matter differently:

AI-Optimized Servers. This is the fastest-growing category in Dell's history and currently the fastest-growing product category in the entire server industry. These are purpose-built systems designed to run GPU-accelerated AI workloads: model training, fine-tuning, and inference at scale. The flagship platforms include the PowerEdge XE9680 (an 8-GPU air-cooled system supporting up to 8 NVIDIA H100, H200, or B200 GPUs, or AMD Instinct MI300X), the PowerEdge XE9712 (NVIDIA GB200 NVL72-based rack-scale system connecting 36 Grace CPUs with 72 Blackwell GPUs), and the XE9780/XE9780L (scaling up to 256 NVIDIA HGX B300 GPUs per rack). These are not single servers - they are factory-configured, pre-cabled, liquid-cooled systems that ship as complete units and integrate Dell's own rear-door heat exchangers and power distribution infrastructure. The liquid-cooled GB200 NVL72 system is up to 25 times more energy-efficient than air-cooled H100 systems, which matters enormously when a hyperscale data center operator is paying tens of millions of dollars per year in electricity.

In FY2026, Dell shipped $25.2 billion worth of AI servers, more than 8 times the prior year. The company exited FY2026 with a $43 billion backlog of AI server orders. Management guided to $50 billion in AI server revenue for FY2027, implying the business will roughly double again. At this pace, Dell's AI server business alone would be one of the largest technology product lines ever built.

Traditional Servers and Networking. The non-AI server business - the PowerEdge R-series and T-series rack and tower servers used for general-purpose computing, virtualization, databases, ERP, and enterprise applications - remains a large and growing business. Demand in this category grew double-digits through FY2026, driven partly by enterprises upgrading aging infrastructure (Dell notes that more than 70% of the installed server base is still running 14th-generation or older PowerEdge servers) and partly by the broader enterprise adoption of AI workloads that require substantial supporting compute. Traditional server revenue was $5.9 billion in Q4 FY26, up 27% year-over-year.

Storage. The storage business carries Dell's deepest intellectual property and historically its highest margins. The portfolio spans six distinct architectures:

• PowerStore is an all-flash storage array designed around a data-centric architecture that allows it to consolidate mixed workloads (block, file, and VMware environments) on a single platform. It has delivered seven consecutive quarters of double-digit demand growth as of Q3 FY26.
• PowerMax is the high-end, mission-critical storage platform for workloads requiring extreme performance and six-nines availability - financial trading, healthcare records, and airline reservation systems. It competes directly with IBM's DS8000 and HPE's Primera/Alletra.
• PowerScale is a scale-out NAS platform (descended from the Isilon acquisition) that handles massive unstructured data workloads - genomics, seismic data, and increasingly AI training datasets and inference checkpoint storage. Dell's ObjectScale.Next claims up to 50% greater throughput per node versus competing solutions.
• ObjectScale is Dell's cloud-native object storage platform for AI/ML data lakes and large-scale archival. The XF960 node achieves 8x greater density than its predecessor.
• PowerFlex (formerly ScaleIO) is a software-defined storage platform for converged and hyper-converged infrastructure.
• PowerProtect is the data protection and backup platform.

Storage revenue was $4.8 billion in Q4 FY26, up 2% overall, but the proprietary Dell IP sub-portfolio (PowerMax, PowerStore, PowerScale, ObjectScale) grew double-digits. The gap between IP storage growth and overall storage growth reflects the mix shift away from legacy hardware toward Dell's modern platforms, which carry better margins.

Storage sits at an interesting intersection: AI training clusters require not just GPU servers but enormous quantities of fast, high-capacity storage for training data, checkpoint saves, and inference KV caches. Dell's storage business is quietly becoming an AI data infrastructure play in addition to its traditional enterprise market.

ISG generated approximately $14.1 billion in revenue in Q3 FY26 (up 24%) and $19.6 billion in Q4 FY26 (up 73%). As a share of total Dell revenue, ISG has gone from roughly 40% historically to over 58% in FY2026.

ISG's core capability: Dell spent decades and billions of dollars building the supply chain relationships, manufacturing processes, and service infrastructure to assemble and deploy server and storage systems at global scale. The company's 25,000+ patent portfolio, $3.1 billion annual R&D budget, and relationships with every major silicon vendor (NVIDIA, Intel, AMD, Broadcom, Marvell, Samsung, Micron) are not easily replicated. What took 40 years to build cannot be rebuilt in 4.

Client Solutions Group (CSG)

CSG is the PC business. It makes and sells laptops, desktops, workstations, and peripherals to commercial and consumer customers. Under the commercial umbrella, Dell sells Latitude laptops (for business users), OptiPlex desktops (office workstations), Precision workstations (for engineers, architects, video editors, and AI developers who need GPU-accelerated local compute), and Vostro products (for small businesses). For consumers and enthusiasts, it sells XPS premium laptops and desktops, and Alienware gaming systems.

CSG commercial revenue has been the more interesting story: six consecutive quarters of year-over-year growth through Q4 FY26, with Dell gaining 100 basis points of market share in Q4 in a market that itself grew 10%. The driver is the Windows 11 transition: Microsoft ended mainstream support for Windows 10 in October 2025. As of Q3 FY26, Dell estimated that over 1 billion of the world's 1.5 billion installed PCs had not yet transitioned to Windows 11. That is a replacement cycle that will play out over several years.

Layered on top of the OS refresh is a hardware transition: AI PCs. Dell launched the first Copilot+ AI PCs in March 2025, and the Dell Pro Max notebooks support discrete neural processing units (NPUs) capable of running a 109-billion parameter model locally. As enterprises evaluate on-device AI - reasoning assistants, local inference, edge AI applications - the incentive to replace aging hardware accelerates.

CSG consumer has been weaker: consumer revenue declined 7% in Q2 FY26 and 19% in Q1 FY26, as post-COVID demand normalization persisted. Dell has chosen not to chase consumer volume at the expense of margin. The commercial market is where Dell wins.

CSG contributed $12.5-13.5 billion per quarter through FY2026. Operating margins in CSG are thinner than ISG - approximately 4-5% versus ISG's 14-15% - which reflects the competitive intensity of the PC market. But CSG serves a critical strategic function: it is Dell's primary point of entry into an enterprise account. An IT director who buys Dell laptops for 10,000 employees is a natural buyer of Dell servers and storage. The integrated relationship is difficult to replicate without the full portfolio.

Segment Comparison

3. Products and Business Detail

The AI Server Portfolio: Systems Engineering at Scale

Dell's AI server portfolio represents perhaps the most technically complex manufacturing challenge in the technology hardware industry today.

The PowerEdge XE9680 is the entry point: a 6U air-cooled rack server housing eight high-end GPUs. Dell supports four GPU types in this form factor - NVIDIA HGX H100 (80GB), HGX H200 (141GB), HGX B200, and AMD Instinct MI300X (192GB). The server is powered by dual Intel Xeon Scalable processors (4th or 5th generation, up to 64 cores each), supports up to 4TB of DDR5 ECC memory across 32 DIMM slots, and requires six 3,200W power supplies in a 3+3 redundant configuration - totaling 19,200W of power per server. Fully loaded, the chassis weighs 251 lbs. A typical AI training cluster might contain 512 or 1,024 of these servers. The XE9680 is where Dell got its first large AI orders in 2023-2024 and built its NVIDIA relationship.

The PowerEdge XE9712 is the current flagship. It is not a server - it is a rack-scale system. The XE9712 implements NVIDIA's GB200 NVL72 architecture: 36 NVIDIA Grace Hopper CPUs connected to 72 NVIDIA Blackwell GPUs via NVLink, creating what NVIDIA describes as "one giant GPU." The 72-GPU NVLink domain can act as a single compute unit, enabling 30x faster real-time inference on trillion-parameter models versus prior-generation H100 systems. The system requires direct liquid cooling - there are no fans inside the hot-aisle/cold-aisle; instead, a manifold delivers coolant to each GPU module. Dell integrates rear-door heat exchangers that can reduce cooling energy costs by up to 60%.

Dell was first to ship the GB200 NVL72-based system commercially. Being first to ship each generation is not accidental - it requires engineering teams to work alongside NVIDIA from the early design phase, often 12-18 months before a product publicly launches, to co-design the thermal management, power delivery, and firmware integration. Dell has managed to be first-to-ship across the Hopper (H100), Hopper successor (H200), Blackwell (B200), and Blackwell Ultra (B300) generations.

The XE9780 and XE9780L represent the current extreme: configurations supporting up to 256 NVIDIA HGX B300 GPUs per rack. Announced at Q1 FY26, these platforms were shipping as of mid-FY26 to neocloud customers running the largest-scale AI training jobs.

The forthcoming PowerEdge XE9812 and XE9880L (announced at Dell Technologies World 2026, global availability 2H 2026) are the next-generation rack-scale systems that will support NVIDIA's Vera Rubin architecture.

The Storage Portfolio: The Margin Engine

Dell's storage business is the direct legacy of the $67 billion EMC acquisition. EMC brought the world's most complete enterprise storage portfolio, and Dell has spent the intervening decade modernizing each platform:

PowerStore launched in 2020 as a ground-up redesign of EMC's Unity platform. It runs a real-time, shared-everything clustered OS (based on EMC's DSSD technology) across nodes, allowing online scale-up without downtime. The platform handles block, file, and VMware VVols natively without additional licensing. Seven consecutive quarters of double-digit demand growth (as of Q3 FY26) indicates it is winning displacement deals against Pure Storage's FlashArray and NetApp's AFF.

PowerMax (descended from EMC's Symmetrix and VMAX) is the mainframe-adjacent, mission-critical tier. It uses a proprietary operating environment (ServiceAbility Manager) with up to six-nines availability and supports FICON mainframe connectivity alongside NVMe/FC and iSCSI. Its primary use case is financial core banking, airline reservations, and healthcare systems where even one second of downtime represents millions of dollars in losses.

PowerScale (descended from Isilon) is a scale-out NAS designed for files at petabyte scale. The architecture allows nodes to be added to a live cluster without downtime, scaling from dozens of terabytes to exabytes. AI training datasets (often consisting of billions of files totaling petabytes), genomics libraries, media repositories, and seismic survey data are the sweet spots. PowerScale's integration with NVIDIA GPU storage protocols (GPUDirect Storage) makes it attractive as an AI training data store.

ObjectScale (descended from ECS - Elastic Cloud Storage) is a Kubernetes-native object store that implements the S3 API. It is designed for unstructured data at cloud scale: AI data lakes, backup targets, surveillance archives, and content delivery storage. The next-generation XF960 node delivers up to 50% greater throughput per node versus alternatives from VAST Data and Weka.

Dell Financial Services: The Invisible Competitive Weapon

Dell Financial Services (DFS) is a captive financing arm that originates, collects, and services receivables related to Dell's products. As of January 30, 2026, DFS managed a $14.3 billion global portfolio of financing receivables, having originated $11.9 billion of new financing in FY2026.

DFS is strategically important in ways that quarterly financials obscure. When an enterprise CFO is deciding between Dell and HPE for a $50 million server order, Dell can offer to finance it. The customer gets DeFferred cash outlay; Dell gets a stickier relationship and a predictable revenue stream. For AI clusters costing $100-500 million per deployment, financing is often the difference between winning and losing a deal. No pure-play server competitor (certainly not Supermicro) can offer this.

Geography

Dell operates in more than 170 countries. Americas is the largest region, contributing approximately 65% of ISG revenue and 60% of CSG revenue. EMEA and APJ make up the remainder. Notably, the sovereign AI opportunity is concentrated outside the US: governments in Europe, the Middle East, Japan, India, and Southeast Asia are investing in national AI infrastructure to maintain technological sovereignty and avoid dependence on US cloud providers. Dell has explicitly called out sovereign customers as a growing portion of its AI pipeline.

Manufacturing

Dell does not own silicon fabs. Its manufacturing model is final assembly: components (GPUs, CPUs, memory, drives, networking ASICs) are procured from their respective manufacturers (NVIDIA, Intel, AMD, Samsung, Micron, Broadcom) and assembled in Dell-owned facilities in seven countries, supplemented by contract manufacturers. Critical to the AI server business is that Dell must secure NVIDIA GPU allocations - meaning it must be on NVIDIA's preferred OEM list and maintain the supply chain relationships to receive GPU shipments before competitors. The SemiAnalysis reporting (and management commentary) confirms this: GPU allocation is not merely a procurement challenge, it is a relationship built over years.

4. Customers

Who Buys From Dell

Dell's customer base divides into three distinct categories with different purchasing dynamics:

Enterprise customers (the largest group by revenue): These are Fortune 500 and Global 2000 companies - banks, insurance companies, manufacturers, healthcare systems, retailers, energy companies, and governments. They buy PCs for employees, servers for data centers, storage for business applications, and increasingly AI infrastructure for internal AI programs. The buying decision inside an enterprise typically involves a Chief Information Officer (CIO) or Chief Technology Officer (CTO) at the strategic level and a procurement team at the operational level. Sales cycles for large deals are 6-12 months. Dell has longstanding relationships across virtually every major enterprise in the world, built over 40 years of selling PCs and servers. These relationships are maintained by Dell's direct enterprise sales force - the largest in the industry, with dedicated account teams for large accounts.

Neocloud customers: CoreWeave, Voltage Park, Lambda Labs, Crusoe, Nebius, and similar companies that build GPU clusters at scale and rent capacity. These customers buy hundreds of thousands of GPUs at a time and require a supplier who can reliably deliver verified, rack-scale systems at speed. Dell's supply chain and first-to-ship status with each NVIDIA GPU generation have made it the preferred OEM for this segment. The SemiAnalysis reporting confirms Dell gained sockets at CoreWeave, Tesla, and xAI - the three largest GPU buyers that use OEM supply chains rather than direct ODM relationships.

Sovereign customers: National governments and quasi-governmental entities building domestic AI infrastructure. France's Sovereign AI initiative, UAE's cloud ambitions, Japan's GENIAC program, and India's AI infrastructure buildout are examples. These customers require on-premises AI infrastructure for data sovereignty reasons and often prefer suppliers with whom they have long-standing government relations. Dell's 170-country footprint and government sales expertise are advantages.

Why They Buy From Dell

There are four specific reasons that repeat across customer types:

Supply chain reliability: In a period when NVIDIA GPU allocations are the most constrained component in the entire technology supply chain, Dell's scale and NVIDIA relationship give it preferential access. A neocloud that needs 10,000 GPUs delivered in Q2 cannot afford a supplier that misses the window.

Verified system integration: A GB200 NVL72 rack requires flawless integration across GPUs, CPUs, NVLink switches, power distribution units, cooling manifolds, network cables, and firmware. Dell has done this hundreds of times. A customer who buys only GPUs from NVIDIA and assembles their own system takes on the integration risk. Dell absorbs it.

Global services: Once a 10,000-GPU cluster is deployed in a data center in Singapore, it needs maintenance. Dell has 97,000 employees globally with service capabilities in 170+ countries. Few competitors match this.

Integrated financing: Dell Financial Services can structure multi-year financing arrangements that allow customers to deploy AI infrastructure today without a single massive capex event.

Switching Costs

Switching costs are moderate-to-high in enterprise, moderate in neocloud. For an enterprise that has standardized on Dell servers and storage, switching to HPE requires re-training the IT staff, re-certifying applications, migrating data from Dell storage to NetApp or Pure, and re-negotiating service contracts. The combined switching cost is significant enough that most enterprises stay with incumbent vendors unless they have a compelling reason to change.

For storage specifically, switching costs are highest. Moving petabytes of data from PowerScale to a competitor NAS is a multi-month project. The PowerMax installed base is effectively locked in for mission-critical workloads.

Customer Concentration

Dell does not disclose revenue concentration by customer (no single customer exceeds 10% of revenue). However, the AI server backlog of $43 billion is likely concentrated among a relatively small number of large neocloud and sovereign customers, which represents both the opportunity and a risk: if a major neocloud customer shifts to a different OEM or to direct ODM procurement, it creates a large hole.

5. Competitive Landscape

AI-Optimized Servers: The Critical Battlefield

The AI server market is where Dell's competitive story is being written in real time.

Super Micro Computer (Supermicro) was the first mover in AI servers. As a Taiwanese-American company operating with an ODM-style, design-to-order model, Supermicro was agile enough to support early GPU server configurations when larger OEMs were still in qualification testing. It gained large positions at early neocloud customers. Supermicro's advantages are speed (it can design and ship a new configuration faster than any competitor), customizability, and price (no large direct sales force overhead). Its disadvantages are: limited services infrastructure, a significant accounting restatement crisis in late 2024 that raised governance concerns among enterprise buyers, and an inability to offer financing or global support at Dell's scale. Technology Business Research (TBR) data shows Dell holding approximately 47% of the OEM AI server market in 2025, with Supermicro's share under pressure following the accounting controversy.

Hewlett Packard Enterprise (HPE) is Dell's most direct competitor across both servers and storage. HPE's ProLiant DL/ML server line competes with Dell PowerEdge; HPE Alletra and Primera compete with Dell PowerStore and PowerMax; HPE Cray competes with Dell's HPC offerings. HPE has leaned heavily into its GreenLake as-a-service model, which is differentiated from Dell's approach. HPE has historically been weaker in the hyperscale AI server market (it was slower to develop liquid-cooled NVIDIA GPU platforms), giving Dell an opening. In storage, HPE has made acquisitions (Nimble Storage in 2017, Zerto in 2021) to close gaps, but Dell's combined PowerStore + PowerMax + PowerScale portfolio is generally considered broader. HPE is the second-place general server competitor.

Lenovo is the world's largest PC maker and a major server competitor in Asia, particularly China. Lenovo's ThinkSystem server line competes globally, and Lenovo has been a top-3 server vendor by volume. However, Lenovo's exposure to China creates US government procurement complications, and its AI server positioning has been weaker than Dell's. In PCs, Lenovo is Dell's most important competitor globally.

HP Inc. competes only in CSG (PCs and peripherals). HP has a stronger consumer brand in some markets and a large installed base, but Dell has consistently outperformed HP in commercial PC share.

NetApp and Pure Storage are the primary pure-play storage competitors. NetApp's ONTAP platform is broadly deployed in enterprise NFS environments and competes with PowerScale. Pure Storage's FlashArray competes with PowerStore and is regarded for its simplicity and all-flash performance. Dell's advantage against both is breadth: no pure-play storage vendor covers Dell's range from entry-level to mission-critical, and none has Dell's integration with servers and services.

Barriers to Entry

Three factors make this industry genuinely hard to enter at Dell's scale:

NVIDIA relationship: NVIDIA does not sell to just anyone. Being a preferred OEM requires years of co-engineering work, demonstrated ability to build compliant systems, and sufficient volume to justify NVIDIA's investment in the relationship. A new entrant cannot simply decide to become a major AI server builder tomorrow.

Global services infrastructure: Dell has 97,000 employees, including tens of thousands of service engineers, distributed globally. Building this takes decades and billions of dollars. It is not a capability that can be acquired.

Enterprise sales relationships: Dell has had dedicated account teams at large enterprises for 20+ years. These relationships mean Dell is in every RFP, every refresh cycle, every new initiative. New entrants start from zero in enterprise sales.

Structural Shifts

The most important structural shift is the move toward rack-scale liquid-cooled AI systems. This transition raises the technical bar significantly above what Supermicro can easily match (liquid cooling requires deeper thermal engineering) and plays to Dell's integration and services strengths. The second structural shift is the emergence of sovereign AI as a demand category - governments do not buy from obscure ODM vendors; they buy from established, globally trusted suppliers. Both trends favor Dell relative to Supermicro.

The existential risk to Dell's AI server business would be hyperscalers (Google, Amazon, Microsoft) and large neocloud customers building their own custom silicon and eliminating GPU purchases. Google's TPUs and Amazon's Trainium are early examples. But this risk is limited to the hyperscale layer; enterprise and sovereign customers do not have the resources to build custom silicon, and they represent the fastest-growing part of Dell's AI pipeline.

6. Industry

What Drives Demand

Dell's addressable market is fundamentally driven by global enterprise IT spend and, increasingly, AI infrastructure investment. The dynamics are layered:

AI infrastructure buildout: The single most important demand driver today. IDC projects global AI infrastructure spending to reach $758 billion by 2029, growing from roughly $334 billion in 2025. Gartner estimates AI-optimized server spending alone at $353 billion in 2026, up from $280 billion in 2025, with a 28% five-year CAGR. The worldwide server market reached a record $125.3 billion in Q4 2025 alone, up 52% year-over-year, per IDC. Dell sits at the center of this cycle.

Enterprise IT refresh cycle: Independent of AI, enterprises periodically replace aging infrastructure. More than 70% of Dell's installed server base is 14th-generation PowerEdge or older, representing a multi-billion-dollar refresh opportunity over the next several years.

PC refresh cycle: The end of Windows 10 support in October 2025, combined with the introduction of AI PCs requiring dedicated NPU silicon, creates a replacement cycle for the 1+ billion PCs running Windows 10 on older hardware. Dell estimated in Q3 FY26 that over 1 billion of the world's 1.5 billion PCs had not yet transitioned to Windows 11.

Digital transformation: The broad enterprise migration of applications to hybrid cloud environments, the adoption of data analytics, and the expansion of digital services all require more compute and storage infrastructure.

Market Structure

The server market is an oligopoly of a few large OEMs (Dell, HPE, Lenovo, Inspur in China) plus a tier of faster-moving specialists (Supermicro, Gigabyte). Dell leads by revenue in the enterprise-facing global server market. IDC confirmed Dell's leadership in server market share driven by AI infrastructure demand.

The storage market is more fragmented: Dell leads by revenue share, followed by NetApp, Pure Storage, IBM, and HPE. The high-growth vector within storage (all-flash, NVMe, cloud-native object storage) is where Dell's new platforms are gaining.

The PC market is a three-way contest between Lenovo (global volume leader), HP Inc., and Dell (strong in commercial enterprise).

Cyclicality

The server market is moderately cyclical. Enterprise IT budgets are correlated with corporate earnings, which are correlated with GDP. However, the current AI capex cycle has a different character: it is driven by competitive necessity (companies must invest in AI or lose to competitors) and by the deployment of large pools of private capital (hyperscaler balance sheets, PE-backed neocloud companies), which are less sensitive to the economic cycle than typical corporate IT budgets.

The PC market is more cyclical and was severely disrupted by the COVID demand surge (2020-2021) and the subsequent hangover (2022-2023). The current refresh cycle driven by Windows 10 EOL provides a durable multi-year tailwind that is independent of the macro cycle.

Regulation and Trade

The US-China technology relationship is the most important regulatory variable for Dell. NVIDIA GPU exports to China are subject to Commerce Department export controls (BIS restrictions), which effectively limits Dell's ability to sell AI servers in China. This constrains the TAM but also reduces competition from Chinese vendors in Western markets. Dell's manufacturing of AMD-based servers has been partially relocated to Mexico, per SemiAnalysis reporting on tariff dynamics.

7. Growth Triggers

Sourced directly from the four most recent earnings calls.

• AI server backlog at $43 billion with FY2027 guidance of $50 billion in AI revenue, doubling from FY2026's $25.2 billion shipped. Management pointed to this as the single largest revenue growth vector in Dell's history. (Q4 FY26 concall, Feb 26, 2026)

"We're entering fiscal 2027 with a $43 billion AI backlog, and we're targeting $50 billion in AI revenue this year. The AI opportunity is meaningfully growing and transforming our business." - Jeff Clarke, Q4 FY26 concall

• AI customer base exceeding 4,000 customers across neoclouds, sovereign entities, and enterprises, with the enterprise segment described as "the fastest-growing part of our pipeline." Enterprise AI deployments are transitioning from proof-of-concept to production, which historically creates durable repeat purchasing. (Q4 FY26 concall, Feb 26, 2026)

• Traditional server demand outpacing supply, with pricing adjustments implemented December 10. Supply tightness for traditional PowerEdge servers implies a demand environment where Dell can take price - a margin tailwind. (Q4 FY26 concall, Feb 26, 2026)

• Windows 10 end-of-life (October 2025) expected to drive a multi-year PC refresh cycle, with management estimating over 1 billion of the world's 1.5 billion installed PCs still needing to upgrade. This structural upgrade cycle spans FY2026 and FY2027 and supports CSG commercial revenue. (Q3 FY26 concall, Nov 25, 2025)

"Over 1 billion of 1.5 billion installed PCs remain unconverted to Windows 11. This is a multiyear tailwind for our commercial PC business."

• Dell was first-to-ship NVIDIA GB200 NVL72 and GB300 solutions commercially. First-to-ship status in each GPU generation creates a competitive advantage in securing early customer orders and establishing the supply chain before competitors. (Q2 FY26 concall, Aug 28, 2025)

• Five-quarter AI pipeline "continues to grow sequentially" and is described as "multiples of our backlog." This implies demand visibility extending well into FY2027 and FY2028. (Q1 FY26 concall, May 29, 2025)

"Our five-quarter pipeline continued to grow sequentially and remains multiples of our backlog." - Jeff Clarke, Q1 FY26 concall

• Sovereign AI as a new and accelerating demand category. Governments in Europe, the Middle East, and Asia-Pacific are funding domestic AI infrastructure projects. This is incremental to the enterprise and neocloud demand streams. (Multiple concalls; specifically called out in Q3 FY26 and Q4 FY26)

• Storage services attach to AI systems: management described current attach rates as "upside and opportunity," implying that the current AI server revenue does not yet include the full potential of PowerScale and ObjectScale storage layered on top of compute clusters. (Q1 FY26 concall, May 29, 2025)

• ISG price increases implemented December 10 (ISG) and January 6 (CSG) to address component cost inflation, with management stating they shortened quote validity periods to better manage cost fluctuations. These pricing actions should protect gross margin rate as input costs rise. (Q4 FY26 concall, Feb 26, 2026)

• AI revenue expected to double again in FY2027, from $25.2 billion to $50 billion, supported by Q1 FY27 guidance of ~$13 billion in AI server revenue in a single quarter. This is consistent with the backlog conversion timing and represents an acceleration in the pace of shipment. (Q4 FY26 concall, Feb 26, 2026)

• Operating leverage: OpEx grew only 5% in Q4 FY26 despite 39% revenue growth, demonstrating that Dell's cost structure does not scale linearly with AI revenue. Management expects continued OpEx efficiency through process modernization. (Q4 FY26 concall, Feb 26, 2026)

8. Key Risks

1. Component Cost Inflation Compressing Margins

Mechanism: DRAM and NAND flash prices - which make up roughly 35% of a server's bill of materials and a comparable share of a PC's BOM - surged aggressively in the second half of FY2026. Spot DRAM hit $2.39 per gigabit (up 5.5 times in six months per Q4 FY26 data), and NAND approached $0.20 per gigabyte (up 4 times). Memory suppliers (Samsung, Micron, SK Hynix) allocate aggressively during AI-driven demand spikes. If Dell's pricing increases (implemented December 2025 and January 2026) do not fully recover these input costs, gross margins compress. Morgan Stanley models have estimated 150 basis points of potential compression if DRAM inflation persists.

Calibration: High probability, moderate magnitude. Dell has demonstrated it can reprice quickly (shortened quote validity, direct-model pricing flexibility), but there is always a lag between cost increases and full recovery, particularly in CSG where retail list prices cannot be changed as dynamically. Jeff Clarke's comment that "the supply environment is as tight as we have ever seen, and input costs are moving higher" (Q4 FY26) indicates this is an acute, near-term concern.

2. AI Server Margin Dilution

Mechanism: AI servers (particularly GPU-dense rack-scale systems) carry lower gross margin rates than traditional servers and dramatically lower margin rates than storage. Dell's stated position is that AI servers are "rate-dilutive but dollar-accretive" - the volume is so large that total gross profit dollars grow even as the percentage shrinks. But if the mix continues shifting overwhelmingly toward AI servers, the overall ISG margin rate could deteriorate even as revenue surges. Management guided mid-teens operating margin for ISG in Q3 FY26, which is achievable but requires everything else in ISG to carry its weight.

Calibration: Structural, ongoing. Dell is aware and managing through value engineering (reducing per-unit cost of AI server assembly), enterprise mix (enterprise customers pay more than neocloud customers for the same hardware), and services attach. But this is a genuine tension between growth rate and margin rate.

3. Demand Timing Non-linearity

Mechanism: AI infrastructure deployment is dependent on inputs that Dell does not control: data center construction, power provisioning, liquid cooling installation, and networking infrastructure. A customer who orders a 10,000-GPU cluster in Q1 may not be ready to receive it until Q3 because their data center is under construction. This creates timing uncertainty in revenue recognition that can cause individual quarters to disappoint even when the multi-year demand picture is intact.

Calibration: Recurring quarter-to-quarter risk. Dell has explicitly called this out in every concall. The $43 billion backlog provides visibility, but the timing of conversion is lumpy.

4. Technology Generation Risk

Mechanism: NVIDIA releases a new GPU generation roughly every 12-18 months. Each transition creates a potential pause in customer orders as buyers wait for the next generation. The Hopper-to-Blackwell transition created some pause in 2024. The Blackwell-to-Vera Rubin transition (expected 2026-2027) may create similar hesitation. Customers who delay orders by one quarter to wait for the next generation can create a meaningful revenue air pocket.

Calibration: Periodic, manageable. Dell's backlog and the urgency of AI deployment timelines reduce the risk of prolonged pauses, but individual quarters during generation transitions can disappoint.

5. Customer Concentration in AI and ODM Risk

Mechanism: A significant share of Dell's AI server backlog is concentrated among a small number of very large neocloud customers. These customers are sophisticated buyers who actively evaluate procurement alternatives: buying directly from ODMs like Foxconn, Quanta, or Wistron, or building custom silicon (like Google with TPUs). If a major neocloud customer decides to bypass Dell and source hardware directly from an ODM at lower cost, it removes a large revenue stream with no equivalent replacement.

Calibration: Moderate probability, high magnitude for specific quarters if it happens. The fact that Dell gained sockets at CoreWeave, Tesla, and xAI (per SemiAnalysis) shows it is winning head-to-head today, but neocloud customers are not loyal. Enterprise customers are stickier.

6. PC Market Cyclicality and Competitive Pressure in CSG

Mechanism: The PC market is cyclical, consumer is soft, and competition from Lenovo and HP Inc. is intense. Dell's consumer business (Alienware, XPS, consumer Inspiron) has been declining. If the Windows 11 refresh cycle is slower than expected - perhaps because enterprise IT budgets face macro pressure - CSG commercial growth could stall.

Calibration: Moderate probability, moderate magnitude. CSG is not the growth story; ISG is. But a significant CSG miss would affect the total company and signal weakness in the enterprise relationship model.

7. Debt Load and Interest Expense

Mechanism: The $67 billion EMC acquisition was substantially debt-financed. Dell still carries significant long-term debt, generating substantial interest expense that dampens GAAP EPS relative to operating performance. Rising interest rates increase the cost of refinancing. Dell Financial Services' $14.3 billion financing portfolio is also sensitive to credit quality deterioration.

Calibration: Manageable given current cash flow generation of $11.2 billion in FY2026, but the debt is a structural feature of the business that limits financial flexibility.

8. Tariff and Trade Policy

Mechanism: US-China trade tensions have already affected NVIDIA's ability to sell high-performance GPUs in China (export controls). If tariffs or export controls expand, they could restrict Dell's supply chain (many components are manufactured in Asia) or reduce its addressable market. Dell has responded by partially moving AMD-based server manufacturing to Mexico, but GPU-centric AI server supply chains remain dependent on Asian manufacturing ecosystems (TSMC in Taiwan, Samsung/Hynix in Korea, Foxconn assembly in Mexico and elsewhere).

Calibration: Geopolitical risk, non-linear. The current US administration's approach to tariffs and export controls is unpredictable. Dell has navigated the existing tariff environment "without customer price increases" (Q2 FY26), but a significant escalation could be disruptive.

9. Walk the Talk

Four concalls used: Q1 FY26 (May 29, 2025), Q2 FY26 (Aug 28, 2025), Q3 FY26 (Nov 25, 2025), Q4 FY26 (Feb 26, 2026).

Dell's management team - co-CEOs Jeff Clarke and Michael Dell, CFO Yvonne McGill (replaced by David Kennedy during FY26), and Chief Customer Officer William Scannell - has demonstrated an unusual and consistent pattern across these four quarters: they have systematically under-guided and over-delivered, particularly on AI. Understanding why this has happened, and whether it reflects honest conservatism or sandbagging, is essential to evaluating the credibility of the FY2027 guidance.

Q1 FY26: Setting the Baseline (May 29, 2025)

Entering Q1 FY26, Dell set full-year FY2026 revenue guidance at $101-105 billion (midpoint $103 billion). For AI servers specifically, management guided to "over $15 billion in AI server shipments" for the full year.

"We're on the plus side of $15 billion for fiscal 2026 AI server revenue. The demand is real, the pipeline is growing, but the timing is non-linear." - Jeff Clarke, Q1 FY26 concall

The under-delivery context: Q1 FY26 actual EPS of $1.55 missed consensus of $1.68 by $0.13. Management attributed this to the non-linear timing of AI demand (customers not yet ready to receive servers) and competitive pricing pressure in CSG. The miss was acknowledged without hedging.

Q2 FY26: The First Upgrade (Aug 28, 2025)

By Q2, it was clear $15 billion was already too low. Dell raised its AI server shipment guidance to $20 billion for the full year, a $5 billion increase in a single quarter. Full-year revenue guidance was also raised from $103 billion to $107 billion (midpoint).

"First-half AI shipments exceeded all of FY2025. We're raising our full-year AI guidance to $20 billion." - Jeff Clarke, Q2 FY26 concall

Key metric: Dell shipped $8.2 billion in AI servers in Q2 alone, meaning in one quarter it had already shipped more than half its original full-year guidance. Management also disclosed "first to ship" status on NVIDIA GB200 and GB300, a competitive signal that resonated strongly with AI-specialist investors.

Q3 FY26: The Second Upgrade (Nov 25, 2025)

Q3 produced another guidance raise. Full-year revenue guidance moved up to $111.7 billion. Management guided Q4 at $31-32 billion and provided what turned out to be conservative AI shipment guidance of "roughly $9.4 billion" for the quarter. The component cost pressure became a central management theme: DRAM and NAND had inflected sharply upward, requiring pricing responses.

Management's mid-quarter price increase announcement (ISG effective December 10) was unusual - companies rarely change prices intra-quarter for enterprise products. This transparency was a credibility signal. Rather than waiting for margin damage to show up in results, management announced the response.

For FY2027, Clarke introduced a floor:

"Mid-teens EPS growth is our starting point for fiscal 2027. That's the floor, not the ceiling." - Jeff Clarke, Q3 FY26 concall

Q4 FY26: The Outcome (Feb 26, 2026)

Q4 FY26 delivered $33.4 billion in revenue against guidance of $31-32 billion - above the top of the range. AI shipments of $9.5 billion were slightly above the $9.4 billion guidance. Full-year AI shipments totaled $25.2 billion against the $15 billion guidance set at the start of the year - a 68% beat versus initial guidance, and a $5 billion beat versus the last-raised guidance of $20 billion.

Full-year revenue of $113.5 billion compared to the initial guidance of $101-105 billion. EPS of $10.30 for the year compared to the initial guide of $9.40. These are not small beats; these are systematic underestimations of a business in the early exponential phase of its AI growth curve.

Credibility Assessment

Management's consistent under-guidance appears to reflect genuine uncertainty about demand timing (the non-linear nature of AI infrastructure deployment) rather than deliberate sandbagging for optics. Clarke's acknowledgment in Q1 that "timing is non-linear" and his willingness to set a specific $15 billion floor rather than an aspirational high number suggests honest forecasting under uncertainty. The team has been transparent about headwinds (component costs, CSG weakness) and has consistently delivered above its own guidance.

The one area where management credibility is harder to assess: the FY2027 guidance of $140 billion and $50 billion in AI revenue is substantially larger than anything Dell has previously managed. The jump from $25.2 billion in AI shipments to $50 billion in a single year assumes the $43 billion backlog converts at pace and that new orders continue to outrun conversion. If the backlog conversion slows (due to data center readiness issues, technology transition pauses, or macro softness), the $50 billion target could prove too aggressive - the one scenario where management's record of conservative guidance could break.

Overall verdict: This is a management team that consistently delivers above its guidance and addresses headwinds promptly and transparently. The four-quarter pattern of under-guide and over-deliver, combined with explicit risk disclosure (component costs, demand timing, CSG weakness), earns high credibility marks.

10. Shareholder Friendliness Index

Dividends: Dell initiated its dividend in fiscal 2023. The progression since then shows exceptional consistency and generosity. In FY2024, the annual dividend was $1.78 per share, a 20% increase over the prior year. In FY2025, it was raised 18% to approximately $2.10 per share. In FY2026, it was raised another 20% to $2.52 per share. Dell has publicly committed to a minimum 10% annual dividend growth rate through fiscal year 2030. Four consecutive years of double-digit increases, starting from a modest initial dividend, indicates management is serious about cash returns and willing to make binding commitments.

Buybacks and Dilution: Dell executed $7.5 billion in total capital returns in FY2026, including 54 million shares repurchased. The company has returned over $18 billion to shareholders since initiating its capital allocation framework in early FY2023, split between $13.6 billion in repurchases and $4.8 billion in dividends. The share count reduction is visible: diluted weighted-average shares fell from 736 million in FY2024 to 720 million in FY2025 to 684 million in FY2026 - a reduction of approximately 52 million shares (7%) over two years. At the Q4 FY26 call, the board approved a new $10 billion incremental share repurchase authorization, bringing total available buyback capacity to a substantial level. Dell's stated policy is to return more than 80% of adjusted free cash flow to shareholders via buybacks and dividends.

Verdict: Dell is a strong capital returner. The combination of growing dividends, an aggressive buyback program, and explicit multi-year capital return commitments puts it firmly in the Returns Capital category. The 7% diluted share count reduction in two years is among the better buyback track records in large-cap technology.

11. Insider Activities

Source: SEC Form 4 filings via EDGAR and secform4.com, covering May 2025 - May 2026.

Buys: No meaningful open-market purchase activity has been identified for Dell insiders in the last 12 months. The absence of insider buying at a stock that has moved significantly upward on AI momentum is notable. There are no cluster-buy signals here.

Analyzing the Sells:

Silver Lake Partners: By far the dominant seller. Silver Lake (through SL SPV-2, L.P., Silver Lake Partners IV, L.P., and Silver Lake Partners V DE (AIV), L.P.) holds Class B shares in Dell, which it must convert to Class C shares before selling on the open market. The pattern of Class B-to-Class C conversion followed by open-market sale has been ongoing since Silver Lake's original investment in the 2013 Dell take-private. This is a private equity firm methodically reducing its position as part of a long-term exit from an investment it made 12+ years ago. Silver Lake still holds approximately 19.3 million Class B shares as of March 2026, meaning these sales represent position reduction, not a complete exit. The reason is standard PE lifecycle: returning capital to fund LPs after a successful long-term investment, not a negative signal about Dell's business outlook. (Form 4 filings, multiple dates March-April 2026)

Jeffrey Clarke: The April 15, 2026 sale of 116,000 shares at $182.48 (~$21.2M) warrants attention. Clarke, as co-CEO and Vice Chairman, has the deepest operational knowledge of the business. The sale price of $182.48 is near the 52-week high at the time of filing. The absence of a disclosed 10b5-1 plan in available filings is a note of caution - though many executives set up plans without filing disclosures until after execution. The size of the sale ($21.2M) is large in absolute terms but represents a modest fraction of Clarke's total compensation and equity stake at a company valued at over $100 billion. Without a confirmed 10b5-1 plan on record, this is worth monitoring but does not rise to a red flag.

David Kennedy (CFO) and William Scannell (CCO): These are smaller executive sells consistent with compensation-related diversification. CFO sells frequently for liquidity and tax reasons, particularly after a strong run-up in the stock. These are standard.

Ellen Kullman (Director): The March 6 sale of 150,346 shares at $145.13 is a material sell by a board member. Given the timing (shortly after Q3 FY26 results were released with strong AI backlog data), this likely reflects planned portfolio management rather than insider concern. Director-level 10b5-1 plans are common.

Net Assessment: The 12-month insider picture is dominated by Silver Lake's ongoing PE exit and by executive diversification sales following a strong multi-year stock appreciation. There are no open-market insider buys, which is a neutral-to-mild negative signal: insiders are not putting personal capital to work alongside shareholders. The Silver Lake selling is clearly mechanical and should not be read as bearish on the business. The Clarke sale deserves monitoring. Overall: neutral to mildly cautious, with no red flags.

12. Scenarios

Bull Case

The core bull thesis is that Dell has built the only enterprise-grade, globally serviced, full-stack AI infrastructure business at scale, and the world is just beginning to deploy AI seriously.

In the bull scenario, the $43 billion backlog converts quickly through FY2027 and is replaced by an even larger wave of enterprise AI orders. The enterprise segment - which management identified as the fastest-growing part of its AI pipeline - proves to be a durable, sticky demand driver: unlike neoclouds that make one large purchase and then pause, enterprise customers deploy in waves as they move from proof-of-concept to production, adding storage, networking, and services with each expansion. Dell's storage business accelerates: as AI training clusters require petabyte-scale, high-throughput storage, PowerScale and ObjectScale become standard components of the AI Factory deployment. The storage attach rate - which management called "below potential" in Q1 FY26 - normalizes upward, driving ISG margin rates higher.

On the PC side, the Windows 11 refresh cycle delivers broad commercial PC replacements through FY2027, aided by AI PC hardware requirements driving an accelerated hardware upgrade cycle that would not have occurred without the NPU transition. Dell captures 100+ basis points of market share per quarter in commercial PCs.

Component cost pressures prove temporary - memory prices normalize as DRAM and NAND capacity expansions (Samsung, Micron, SK Hynix have all announced significant new fab investments) bring supply into balance. Price increases executed in late 2025 and early 2026 prove durable as demand outpaces available supply, so Dell retains the benefit of higher ASPs even as its own costs moderate. ISG operating margins trend toward the high teens.

By FY2028, Dell has a legitimate claim to $60-70 billion in AI infrastructure revenue, owns the enterprise AI deployment market, and generates $15+ billion in annual free cash flow. The buyback program at current authorization levels retires 10%+ of the share count, compounding EPS growth beyond the underlying business growth.

Base Case

In the base case, Dell delivers on its $140 billion FY2027 revenue guidance and roughly $50 billion in AI server revenue, but the execution is not smooth. Component cost inflation proves stickier than management expects through the first half of FY2027, compressing ISG gross margins by 50-100 basis points versus the FY2026 exit rate. Price increases partially compensate but do not fully cover.

The AI backlog converts as expected, but the five-quarter pipeline growth rate decelerates as the early neocloud buildout matures. Enterprise AI demand fills in the gap but at a slower pace than neoclouds - enterprise sales cycles are longer, proof-of-concept deployments take 6-9 months to convert to production. CSG commercial PCs grow mid-single digits on Windows 11 refresh, but consumer remains soft.

Dell hits its mid-teens EPS growth target for FY2027 and increases the dividend again, maintaining its commitment to 10%+ annual dividend growth. The buyback continues at a meaningful pace. Share count falls by another 4-5% over the year. The business is healthy and growing, but the multiple re-rating that the AI tailwind has driven starts to moderate as the market recalibrates to a more normalized (if still impressive) growth rate.

The most likely path from here is a business that is genuinely larger in two years than anyone expected it to be in FY2024, with AI as a structurally higher revenue and profit contributor than the old server business ever was.

Bear Case

The bear case does not require the AI boom to end - it requires a few specific things to go wrong simultaneously.

First: one or two major neocloud customers (which account for a disproportionate share of Dell's AI backlog) decide to bypass Dell in favor of direct ODM procurement, citing cost pressure. The $43 billion backlog turns out to be softer than advertised, with customers cancelling or deferring a significant portion. Dell has no contractual protection against cancellation - these are orders, not signed take-or-pay contracts.

Second: memory cost inflation runs hotter than Dell's pricing actions can recover, compressing gross margins by 200+ basis points in ISG. Combined with the AI server mix shift (which is already dilutive to margin rate), ISG operating margins dip from 14-15% toward 10%, eliminating the dollar-contribution benefit of the volume growth.

Third: the Windows 11 refresh cycle proves slower than estimated. Enterprise IT budgets face macro pressure in a slowing economy. CSG commercial growth stalls at 0-2%, and consumer continues to decline. The two-segment picture darkens together.

In this scenario, FY2027 revenue comes in at $125-130 billion (below the $140 billion guidance), and EPS grows single-digits or misses FY2026 levels entirely. The stock de-rates. The company's debt load becomes a more prominent concern as free cash flow narrows. Silver Lake's continued selling creates an overhang.

The bear case is not impossible - neocloud customer concentration and component cost volatility are real - but it requires multiple things to go wrong at the same time and ignores the structural tailwinds of enterprise AI adoption, sovereign AI, and the PC refresh cycle.

13. Further Reading

• How Dell Is Beating Supermicro - SemiAnalysis, 2024-05-30 [paid]

Sources:

• Dell Technologies Q4 FY2026 Earnings Call Transcript - The Motley Fool
• Dell Technologies Q4 FY2026 Full-Year Results Press Release
• Dell Technologies Q3 FY2026 Earnings Call Transcript - Seeking Alpha
• Dell Technologies Q2 FY2026 Earnings Call - MarketBeat
• Dell Technologies Q1 FY2026 Earnings Call - MarketBeat
• Dell Technologies Form 10-K FY2026 - SEC EDGAR
• Dell Technologies AI Server Market Leadership - InfotechLead
• IDC Worldwide Server Market Record 2025
• IDC AI Infrastructure Spending to Reach $758Bn by 2029
• Dell Technologies Insider Transactions - secform4.com
• Silver Lake Form 4 Filings - StockTitan
• Dell PowerEdge XE9712 GB200 NVL72 Details - Tweaktown
• Dell Technologies World 2026 - SiliconAngle
• How Dell Is Beating Supermicro - SemiAnalysis (paywalled)
• Dell AI Factory with NVIDIA Press Release
• Dell PowerEdge Named 2025 Market and Innovation Leader in Servers for AI