Dell Technologies

Phase A — Understand the business

Company Overview

Dell is the world's largest "full-stack" IT hardware OEM: it designs, manufactures, sells, and services PCs and data-center infrastructure (servers, storage, networking), wrapped in financing (Dell Financial Services) and support/deployment services. Two reportable segments:

• Infrastructure Solutions Group (ISG) — servers (now split into AI-optimized and traditional), networking, and storage. FY2026 revenue $60,826M, +40% YoY; operating income $7,111M (11.7% margin).
• Client Solutions Group (CSG) — commercial + consumer PCs and peripherals. FY2026 revenue $50,984M, +5% YoY; operating income $2,833M (5.6% margin).

Consolidated FY2026: revenue $113,538M (+19%), GAAP operating income $8,149M, net income $5,936M, diluted EPS $8.68, non-GAAP EPS $10.30. Products are 79.6% of revenue, services 20.4%.

Business model in plain terms. Dell is a volume-and-velocity machine, not a margin machine. It wins by being the lowest-cost, highest-scale assembler/integrator/servicer of other people's silicon (Intel/AMD CPUs, NVIDIA GPUs, Micron/Samsung/SK Hynix memory, Seagate/WD drives). The differentiation is the supply chain, the direct go-to-market, the global services footprint, and the balance sheet — not the components. The recurring, high-margin layer is services (44.8% gross margin in FY2026 vs 13.7% on products) and deferred revenue ($26.9B at end of FY2026).

Contract structure / payment terms. Mostly transactional hardware with attached multi-year support; remaining performance obligations ~$97B as of May 1, 2026, ~80% recognized within 12 months. The AI-server book is a mix of hyperscaler/neocloud strategic deals (large, concentrated, aggressively priced) and enterprise. Backlog is the single most important non-GAAP disclosure: ~$51.3B AI-related backlog reported alongside Q1 FY27.

Supply Chain

Dell sits in the middle of the AI build-out chain — it is an integrator, so it is structurally exposed both upstream (component scarcity/inflation) and downstream (customer concentration). Named stakeholders:

Upstream (suppliers / inputs):

• GPUs: NVIDIA (the gating input for AI-optimized servers — GB200/GB300 NVL72 rack-scale). AMD (MI300-class) as secondary.
• CPUs: Intel, AMD.
• Memory/storage: Micron, Samsung, SK Hynix (DRAM/HBM/NAND); Seagate, Western Digital (HDD). This is the inflation chokepoint — see Lens 5/10.
• Networking/optics, substrates, power/thermal (liquid cooling) for rack-scale designs.
• Contract manufacturing / Dell's own factories: Dell retooled its own plants (Round Rock and global) for rack-scale integration in 2024–25.

Downstream (customers / channel):

• Hyperscalers + neoclouds + AI labs: the AI-server demand engine. Named: xAI ("Colossus" Memphis cluster built partly on Dell; advanced talks for a reported ~$5B Dell server order); CoreWeave-type neoclouds; large enterprises standing up private AI.
• Enterprise + government + commercial PC base (CSG).
• Channel/distribution + direct (Dell.com, direct sales force).

Chokepoints / single-source dependencies:

1. NVIDIA GPU allocation — the hard gate on AI-server revenue timing; "non-linearity in the timing of demand and subsequent shipments" is management's own phrase.
2. Memory (DRAM/NAND) — supply and price. Mgmt: "the cost basis is going up across all products… we're repricing, it feels like, every day". DRAM spot reportedly +5.5x over six months.
3. Dell deliberately runs minimal component/product inventory — efficient in steady state, fragile in a supply shock. (Inventory nonetheless climbed to $10.4B at FY26-end as it pre-bought AI components — see Lens 10.)

Names or it didn't happen: NVIDIA → Dell → xAI/neoclouds is the load-bearing AI artery; Micron/Samsung/SK Hynix → Dell is the margin-determining one.

Competitive Advantages (moats)

Dell's moat is operational, not product — and it is real but shallow per-unit in AI:

• Scale + supply-chain mastery (the core moat). Dell is the #1 server vendor by revenue and ships ~a fifth of all AI-optimized servers globally. Procurement leverage on NVIDIA allocation, memory, and logistics is a genuine edge over smaller integrators. Jeff Clarke is widely credited as the architect of this.
• Direct + services + financing (switching costs). Multi-year support contracts, DFS financing, and the installed base create stickiness on the enterprise side. RPO ~$97B and deferred revenue $26.9B quantify it.
• Brand + trust at enterprise procurement. In a world where SMCI is under a DOJ/governance cloud (see Lens 10/13), "nobody gets fired for buying Dell" is worth basis points of share.
• Balance-sheet moat. Dell can finance enormous AI working-capital swings (AP +$12.7B in FY26) that a smaller rival cannot — see Lens 5.

Bargaining power — who needs whom:

• vs NVIDIA: weak. NVIDIA holds the scarce input and the pricing power; Dell is one of several integrators competing for allocation. Dell needs NVIDIA more than the reverse.
• vs memory suppliers: weak right now (shortage → suppliers have pricing power).
• vs customers: mixed. Strong vs fragmented enterprise; weak vs hyperscalers/AI labs, who are sophisticated, can multi-source (SMCI, HPE, ODMs like Foxconn/Wiwynn), and extract aggressive pricing. This asymmetry is the entire margin story.

Durability verdict: the scale/logistics/services moat is durable and widening as SMCI stumbles. But it does not confer pricing power on the AI-server unit itself — Dell is a toll-taker on a low toll, defended by volume. That is a different (and lower-quality) moat than the market's ~22x multiple implies.

Segments

All figures (FY annual) and (Q1 FY27).

By product segment — FY2026 (ended Jan 30, 2026):

By geography — FY2026: US $63,140M (+24%), Foreign $50,398M (+13%); no single foreign country ≥10%.

The trend and the cause (this IS the thesis):

• AI-optimized servers went $1.9B → $9.3B → $24.7B in three years (FY24→25→26), then $16.1B in Q1 FY27 alone (+757% YoY). This is the fastest-scaling line Dell has ever had.
• But ISG operating margin fell 110bps to 11.7% precisely because of that mix shift — AI servers carry lower gross margin than storage/traditional servers. Decelerating margin on accelerating revenue is the structural tension.
• Storage (the high-margin ISG anchor) is flat (+1%) — the part of ISG that should defend margin isn't growing.
• CSG (PCs) is a low-growth, margin-eroding cash cow — commercial +8% but consumer −8%, blended +5%, op margin down to 5.6%. The PC refresh (Windows 10 EOL, AI PCs) is a tailwind but not transformative.

Q1 FY27 acceleration: ISG $29,009M (+181% YoY), of which AI servers $16,132M; CSG $14,609M (+17%). ISG operating income $3,055M → 10.5% segment margin (vs 9.7% in Q1 FY26: $998M/$10,317M) — a modest sequential margin recovery despite heavier AI mix, worth watching.

---

Phase B — Measure performance

Earnings Result (latest print: Q1 FY2027, ended May 1, 2026)

The blowout. All `` unless noted.

• Revenue $43,842M, +88% YoY (vs ~$35.4B consensus) — fastest growth since the 2018 re-listing.
• GAAP diluted EPS $5.24 (+282%); non-GAAP diluted EPS $4.86 (+214%) vs ~$2.94 consensus. (GAAP > non-GAAP this quarter because of a +$631M fair-value markup on equity investments, mostly "a single investee" — likely a private AI name marked up ~$0.6B. Strip it: non-GAAP $4.86 is the cleaner number.)
• Lines that drove it: AI-optimized servers $16,132M (+757%); $24.4B AI orders booked in the quarter. ISG $29,009M (+181%).
• Margins: consolidated gross margin compressed (products GM 13.8% vs 14.1% YoY) — AI mix dilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company., exactly as guided. But operating leverage was enormous: GAAP operating income $3,656M (+214%); non-GAAP operating income $4,235M (+154%).
• Guidance RAISED hard (the real catalyst): FY2027 revenue to $165–169B ($167B mid, +47% YoY); AI-optimized server revenue to ~$60B (+144%); non-GAAP diluted EPS raised to ~$17.90 mid (+74%) — up from the ~$12.90 (+25%) guide given in Feb 2026. Note the conflict: several aggregator pieces still quote the stale $12.90 Feb guide as "FY27 EPS" — the current (post-Q1) company guide is $17.90 mid. Use $17.90.
• Balance-sheet flags: record Q1 OCF $4,081M; FCFFree cash flowCash left after paying to run and maintain the business. Unlike profit, it is hard to flatter with accounting choices. $3,118M. Backlog ~$51.3B.
• Market reaction: +32% on May 29, 2026 — Dell's best day ever (narrowly topping the +31.6% of Mar 1, 2024); stock ran from ~$242 (May 20) to >$420. What was priced in: not this. The size of the guidance raise repriced the whole AI-server TAM for Dell.

Unusual vs its own history: an 88% revenue quarter for a $100B+ hardware company is unprecedented; so is a +$27B intra-year revenue-guide raise. The direction (AI up, margin% down, EPS up on leverage) is consistent with FY26 — just violently accelerated.

Earnings Calls (sentiment trend)

No transcripts/ on the shelf → ``-sourced. Across the last ~4 calls (Q2 FY26 → Q1 FY27), management tone has shifted from "AI is a meaningful opportunity" to "AI is transforming the company". Recurring phrases now: "AI Factory," "backlog," "non-linearity of shipments," "disciplined pricing," "world-class supply chain." The new phrase that matters: "the cost basis is going up across all products… we're repricing every day" (Clarke) — management is pre-warning on memory/component inflation. What they've stopped emphasizing: the old "PC recovery" narrative and VMware-resale (terminated Mar 2024). Net sentiment: maximally confident on demand, openly cautious on cost — an honest, bullish-but-hedged posture, not promotional.

Comps

Peer set: AI-server/IT-hardware integrators + the enterprise-infra incumbents. Multiples are `` with date or n/a. Never fabricated.

Stock-Price Catalysts (>5% moves, ~last 5 yrs) — ``

The tape says DELL is now an AI-server beta, and the catalyst is the AI-server guide, full stop:

• May 29, 2026: +32% (best day ever) — Q1 FY27 print + the $167B/$60B/$17.90 guidance raise.
• Mar 1, 2024: +31.6% — the original AI-server breakout quarter (the first time AI orders surprised).
• 2026 YTD: +~234% by late May; 52-wk range $110.22 → $469.47 — a ~4x range in a year.
• Recurring pattern: every >5% move clusters on earnings day and is driven by the AI-server order/backlog number and the forward AI revenue guide — not PCs, not storage, not macro. Secondary movers: NVIDIA product cycle headlines (GB200/GB300 availability) and SMCI governance news (share-shift narrative).

What the market actually reacts to: the trajectory of the AI-server backlog and the forward AI revenue guide. That is the one number that moves DELL. It is also the most estimate-sensitive and the least margin-rich — which is the whole tension.

---

Phase C — Judge people & books

Management

• Track record (strong). Michael Dell — founder (1984), Chairman & CEO; took the company private (2013, with Silver Lake), engineered the $67B EMC acquisition (2016), the VMware tracking-stock unwind, and the 2018 re-listing. A genuine operator-owner with a multi-decade record of value creation and hard pivots. Jeff Clarke — Vice Chairman & COO, the supply-chain/operations architect; ~40 years at Dell. David Kennedy — CFO since Sept 2025 (internal promotion, joined 1998) — note the recent CFO transition into the most consequential guidance period in Dell's history.
• Tenure & skin in the game (very high). Michael Dell owns ~40% of the company and, with Silver Lake (~7.2% economic / ~13% voting), controls it via super-voting Class A/B shares (10 votes each) vs Class C (1 vote). This is a founder-controlled company — aligned on the upside, but minority public holders (Class C) have little governance recourse.
• Capital allocation (shareholder-friendly, aggressive). FY2026: ~$6.0B buybacks (≈54M shares) + ~$1.41B dividends = ~$7.5B returned. Post-FY26: +$10B buyback authorization (→ ~$15.2B remaining) and a 20% dividend hike to $0.63/qtr. Investment-grade rating is an explicit constraint they manage to.
• The capital-structure red flag: Dell carries a negative total stockholders' equity of −$2,470M at FY26-end — a deliberate artifact of years of buybacks ($14.5B treasury stock) funded by debt + the negative-working-capital model, not distress. It means ROE/ROIC are not meaningfully computable (negative-equity denominator) — a real reason the comp table shows ROE "n/a." It also means there is no equity cushion; the model depends on continued cash generation and AP float.
• Red flags (governance). Clarke's $132.4M performance-option grant (May 2026) is large, though performance-vested. Super-voting control + a founder-CEO + a brand-new CFO during a hyper-growth guide is a "trust the operator" setup — fine while execution is flawless, less comfortable if the AI cycle wobbles.
• Archetype: founder-operator (Michael) + lifer-operator (Clarke). For this stage — a hardware scale game in a land-grab — that operational DNA is exactly the right archetype. The risk is the archetype's blind spot: chasing revenue/share at the expense of return on capital.

Forensic Red Flags

Forensic-analyst lens. Dell's accounting is clean (PwC unqualified opinion, ICFR attested ) and there are no SEC enforcement findings (below). The flags here are quality-of-earnings and balance-sheet-fragility, not fraud:

1. Receivables outrunning revenue (the headline flag). Accounts receivable, net jumped +71% to $17,585M (from $10,298M) at FY26-end — faster than revenue (+19%) — "primarily driven by… AI-optimized servers". AR drag was −$7,022M of operating cash flow in FY26. Large AI customers stretch terms; concentration + slower collection is a watch item.
2. The AP-funded OCF engine (the most important forensic point). FY26 OCF of $11,185M (+147%) was manufactured by a +$12,665M swing in accounts payable that more than offset the AR (−$7.0B) and inventory (−$4.0B) drags. AP hit $33,630M (vs receivables $17.6B) — Dell runs a deeply negative cash-conversion cycle (it gets paid before it pays suppliers). This is a structural strength in steady growth, but it is a reversing tailwind: if AI-server growth decelerates or suppliers tighten terms, the AP float unwinds and OCF can fall faster than earnings. OCF quality is lower than the headline +147% suggests.
3. Inventory build. Inventories +55% to $10,437M ($6.7B prior); inventory was a −$3,987M operating cash drain. Management pre-bought AI components ("increased our purchases of certain components… increased inventory levels, higher purchase obligations") — sensible given scarcity, but it pairs with $16.8B of FY27 purchase obligations in a market where component prices are volatile. Markdown risk if AI demand air-pockets.
4. Non-GAAP flattering. Q1 FY27 GAAP EPS ($5.24) exceeded non-GAAP ($4.86) only because of a +$631M equity-investment markup — a non-operating, mark-to-model gain on a private holding (mostly "a single investee"). Cumulative unrealized gains on non-marketable securities reached $1.8B. These marks are real income but low-quality and reversible. SBC ($723M FY26) is added back to non-GAAP as usual.
5. Negative equity / no cushion — see Lens 9. Total debt $31.5B ($7,990M short + $23,513M long), of which core debt $17.0B and DFS-related $14.6B. Manageable against $11B+ OCF, but there is zero book-equity buffer.

Regulatory findings (required sub-section):

• SEC Litigation Releases / AAERs: None. regulatory/regulatory-findings.md (fetched 2026-06-24 via SEC EDGAR EFTS) returns 0 LR and 0 AAER for "Dell Technologies" over 2021-06-24 → 2026-06-24.
• 10-K Item 3 / Note 11 Legal Matters (company's own disclosure): Item 3 incorporates "Legal Matters" by reference. The one named matter: a putative ERISA class action — Lowbruck et al. v. Dell Technologies Inc. (W.D. Tex., filed Jan 28, 2026) — alleging breach of fiduciary duty over Dell 401(k) plan investment options; Dell "intends to vigorously defend." Management states that as of Jan 30, 2026 it does not believe a reasonably possible material loss exceeding accruals exists across all matters. Routine; not thesis-relevant.
• Non-SEC enforcement (web): No material FTC/DOJ/EU enforcement action against Dell surfaced. (The DOJ action in the AI-server space targets Super Micro–linked individuals over ~$510M of illegal NVIDIA-server exports to China — a competitor issue that benefits Dell's share narrative, not a Dell liability).
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-24. The risks here are accounting-quality (AP float, AR, inventory, equity marks), not integrity.

---

Phase D — Project & stress-test

Forward Projection (EPS, next 3 fiscal years)

Built bottom-up from the latest actuals + the current (post-Q1) company guide. Output ``; the FY27 base anchors to management's own raised guide.

FY2027 (ending ~Jan 2027) — base = company guide.

• Revenue $167B mid (+47%); AI-optimized servers ~$60B; non-GAAP diluted EPS ~$17.90 mid.
• Sanity check: Q1 FY27 non-GAAP EPS was $4.86; $17.90 ÷ 4 ≈ $4.48/qtr — i.e., guide implies roughly flat-to-modestly-up sequential quarterly EPS as AI mix dilutes margin even as revenue climbs. Internally consistent.
• Base FY27 non-GAAP EPS: $17.90.

FY2028 — base / bull / bear ``:

• Drivers: AI-server revenue growth decelerates from +144% to a still-large +25–40% as the backlog converts and comps harden; CSG low-single-digit (AI-PC refresh); storage/traditional flattish; operating leverage partly offset by AI-margin dilution + memory-cost inflation; ~2–3% share-count reduction from the $15.2B buyback.
• Base: revenue ~$200B; non-GAAP EPS ~$21.
• Bull: AI servers compound harder (sovereign/enterprise AI broadens, SMCI share bleeds to Dell), memory eases → revenue ~$215B, EPS ~$24.
• Bear: AI capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. digestion air-pocket + memory-cost squeeze on CSG → revenue ~$185B, EPS ~$18 (flat).

FY2029 — base ``: AI growth normalizes to +15–20%; EPS ~$24–25 base, wide bull/bear band ($20 bear / $29 bull) entirely governed by (a) the durability of AI-server demand past the first build-out wave and (b) whether Dell ever recaptures gross margin as the mix matures and services/storage re-grow.

The single sensitivity that dominates all of it: AI-server operating margin. Management has guided AI-server profitability toward mid-single-digit operating margin. At $60B+ of AI revenue, every 100bps of AI op-margin ≈ ~$600M of operating income ≈ ~$0.70 of EPS. The EPS path is far more margin-sensitive than revenue-sensitive from here.

Forecast log: our model create intentionally skipped (per --watchlist rule — log a Brier forecast only on a genuinely committed base case in an interactive our position log pass). Candidate to log when promoted: "DELL FY2027 non-GAAP diluted EPS ≥ $17.90, p≈0.65, resolves 2027-03-15."

Bull vs Bear

Bull case. Dell is the scaled, trusted, financeable prime contractor for the enterprise+neocloud AI build-out at the exact moment its strongest pure-play rival (SMCI) is hobbled by governance/DOJ issues. The AI-server line went $1.9B→$9.3B→$24.7B→$60B-guided in four years; backlog ~$51.3B gives multi-quarter visibility; RPO ~$97B underpins it. Operating leverage is ferocious (op income +154–214% on +88% revenue). Capital returns are large and growing ($7.5B/yr, +$10B buyback, +20% dividend). On the current $17.90 guide the stock is ~23x forward — not cheap, but not absurd for ~40%+ revenue growth, if AI demand persists and margin holds. The contrarian-bull read: the market is still partly anchored to "cyclical box-maker," and a multi-year AI-infrastructure refresh (training → inference → enterprise private AI → sovereign) is a longer runwayRunwayHow long the cash lasts at the current rate of spending. It shortens the moment spending rises, which is why a figure taken from a quiet quarter flatters. than a 22x multiple discounts.

Bear case (2–3 things that could permanently impair the thesis or the multiple):

1. Margin never recovers — the mix is the destiny. AI servers are structurally mid-single-digit-op-margin, NVIDIA owns the value, and hyperscaler customers have the bargaining power. Gross margin already fell 24.3% (FY24) → 20.4% (FY26). If Dell is permanently a low-single-digit-net-margin pass-through on NVIDIA silicon, then a 20%+ forward multiple on that business is the wrong multiple — and the de-rate is the loss.
2. The working-capital engine reverses. OCF was AP-float-funded (+$12.7B AP swing). A demand air-pocket or supplier-term tightening unwinds the float and OCF falls faster than EPS — puncturing the "record cash generation" narrative that justifies the buyback.
3. Memory-cost inflation eats CSG and squeezes ISG. DRAM/NAND up sharply; Clarke is "repricing every day." Dell can't fully pass it through in competitive PC and strategic AI deals → margin compression on both segments at once.

Pre-mortem (it's Dec 2027, the thesis broke — what happened?): AI capex digestion hit; hyperscalers paused/renegotiated the largest orders; backlog converted at lower margin than guided; memory costs stayed high; OCF disappointed as AP normalized; the stock de-rated from ~22x to ~12x even on flat EPS — a ~45% drawdown driven entirely by multiple compression, not an earnings collapse.

Are multiples too high? For the business's quality, yes — at the high end of fair. DELL has re-rated from a ~6–11x cyclical to a ~20x+ "AI compounder." That re-rating is justified only if AI-server demand is durable AND margin stabilizes. Both are plausible; neither is proven.

Contrarian view (what the market refuses to see): the bull crowd is anchored on the revenue trajectory (easy to see, $60B AI) and under-weighting that this is the lowest-margin, most-customer-concentrated, most-NVIDIA-dependent revenue Dell has ever booked — and that the cash generation is partly an accounting-of-timing artifact (AP float). The thing to watch is not the next AI order number; it's AI-server operating margin and the cash-conversion cycle.

Devil's Advocate (short-seller)

Dismantling the bull case.

• Where revenue is concentrated / what breaks it: AI-optimized servers are ~37% of FY26 ISG and rising fast; that revenue depends on (a) NVIDIA allocation Dell doesn't control and (b) a handful of mega-buyers (xAI/neoclouds/hyperscalers) who can multi-source and renegotiate. Concentration shift = a single large customer pausing = a backlog air-pocket = a -30% revenue-growth quarter that the +32% stock cannot absorb at 22x.
• Why the moat is weaker than bulls think: Dell's moat is logistics/scale/services — it confers share, not price. On the AI unit, Dell is a toll-taker on NVIDIA's toll road, competing with HPE and the ODMs (Foxconn/Wiwynn/Quanta) who sell direct to hyperscalers at even thinner margin. The moat doesn't stop margin erosion; it just wins the low-margin volume.
• Most dangerous competitor bulls underestimate: the ODMs / white-box vendors (and a rehabilitated SMCI). If hyperscalers keep shifting AI buying to direct-from-ODM, Dell's AI-server TAM is capped at the enterprise/neocloud slice — smaller and slower than the $60B guide extrapolates.
• Worst capital-allocation optics: buying back $6B of stock into a 200%+ run while carrying negative book equity and rising debt — financially defensible (negative-WC model), but it means there's no balance-sheet cushion if the AI cycle turns; the buyback is pro-cyclical.
• What must hold for today's ~$419 price: AI-server revenue ≥$60B and margin ≥ mid-single-digit and memory inflation contained and AP float intact and the multiple stays ~20x+. That's a stack of "ands."
• If growth disappoints 20–30%: FY28 EPS bear-case ~$18 (flat) on a de-rated ~12–14x → ~$220–250 stock = ~40–45% downside from ~$419. The asymmetry at this price is no longer favorable.
• Single permanent-impairment scenario: AI training capex proves front-loaded (a 2024–26 build-out wave that digests in 2027–28), hyperscalers in-source/ODM the next wave, and Dell is left as a flat-margin enterprise-hardware vendor that the market re-rates back toward HPE's ~9x. Plausibility: medium. Not a base case, but not tail-risk either.

Management Questions (ordered by information value)

1. What is the operating margin on the AI-optimized server business today, and what is the realistic steady-state range once the current land-grab pricing normalizes?
2. Of the ~$51B AI backlog, what share is concentrated in your top 3 customers, and what are the cancellation/repricing terms if AI capex pauses?
3. The +$12.7B FY26 accounts-payable swing funded most of your OCF growth — what happens to operating cash flow if AI-server growth (not level) decelerates and the AP float stops expanding?
4. How much of the ~$60B FY27 AI revenue is firm PO vs. demand signal, and how exposed is the conversion to NVIDIA GB300 availability timing?
5. With DRAM/NAND up multiples, how much memory-cost inflation can you actually pass through in (a) strategic AI deals and (b) competitive commercial PCs — quantify the FY27 margin headwind.
6. Why continue $6B+ buybacks while carrying negative book equity and rising debt, rather than building an equity cushion ahead of a potential AI-capex digestion?
7. What is your gross-margin recovery path — does storage/services/traditional-server mix ever re-rate ISG margin back toward 13%+, or is 11–12% the new ceiling?
8. How real is the SMCI share-shift, and how much of your AI-order acceleration is durable demand vs. a one-time reallocation that ODMs could reclaim?
9. What is the return on the incremental capital (working capital + capex + buyback) you're deploying into the AI cycle — what ROIC are you underwriting?
10. Inventory +55% and $16.8B of FY27 purchase obligations — what's the markdown/obsolescence exposure if AI demand air-pockets or NVIDIA transitions components faster than expected?
11. What is the AI-PC refresh actually worth to CSG in FY27–28 in revenue and margin — or is it a narrative more than a number?
12. How do you defend against hyperscalers in-sourcing / buying direct from ODMs for the next AI wave — what keeps the neocloud/enterprise buyer with Dell?
13. With a new CFO (since Sept 2025) during the largest guidance ramp in company history, what changes (if any) to guidance philosophy or disclosure should investors expect?
14. What would cause you to cut the FY27 AI-server guide, and what leading indicators (order pace, backlog aging, customer concentration) are you watching?
15. Five years out, is Dell a higher-margin AI-systems-and-services company or a larger, lower-margin hardware pass-through — and which are you optimizing the capital structure for?

---