The Zacks Analyst Blog Highlights Dell, Celestica, Lumentum and Vertiv

Zacks highlights AI data-center infrastructure demand with $720B–$745B in 2026 capex expected from Amazon, Alphabet, Microsoft, and Meta. Reported results point to strong momentum: Dell Q1 FY2027 revenue hit $43.84B (+88% Y/Y) with AI server revenue of $16.1B and AI server orders at $24.4B, and it raised FY2027 AI server revenue expectations to ~ $60B; Celestica Q2 2026 revenue rose to $4.70B (+62% Y/Y) and raised FY2026 outlook to $20.5B revenue; Lumentum Q4 FY2026 net revenue reached a record $1.01B (>2x Y/Y) with guidance implying Q1 FY2027 revenue up 130% Y/Y; Vertiv Q2 2026 revenue increased 24% to $3.27B and raised FY2026 guidance to $13.8B–$14.2B. Overall, Zacks frames DELL, CLS, LITE, and VRT as high-visibility beneficiaries of the AI infrastructure buildout (Zacks Ranks #1–#2, Growth Score A).

Analysis

This is less a pure AI demand story than a capex-taxonomy trade: the second wave of beneficiaries are the bottlenecks around racks, optics and thermal management, not the chips themselves. That shifts relative value toward the names with the highest conversion of hyperscaler spending into backlog visibility and the least exposure to end-demand substitution; on that screen, CLS and VRT look cleaner than DELL because they are closer to the infrastructure tollbooth and less dependent on a single product cycle.

The immediate risk is that the market has already crowded into the "AI picks-and-shovels" basket, so good news may be a multiple maintenance event rather than a multiple expansion event. Over 1-3 months, the key catalyst is whether cloud capex commentary from AMZN/MSFT/GOOGL/META confirms that 2026 spending is pulling forward rather than normalizing; if capex growth slows, DELL’s revenue leverage can fade quickly because server demand is easier to defer than power/cooling or optical upgrades. LITE has the highest upside torque, but it is also the most fragile to supply-chain relief, because any easing in component tightness could compress near-term pricing power even if unit demand stays strong.

The contrarian view is that the real winner may be the enablers of power delivery, not the compute OEMs: if data-center density keeps rising, VRT’s growth should prove more durable than server-side names that face inventory and working-capital drag. A second-order loser could be the broader enterprise hardware set, where AI-related mix can cannibalize budget share without improving margins. The thesis is falsified if hyperscaler capex guides down, if DELL backlog conversion disappoints in the next quarter, or if LITE/VRT fail to hold recent estimates despite continued cloud spending.

Key Decisions for Investors

• Long CLS on a 1-3 month horizon as the cleanest operating-leverage expression of AI infrastructure; use a pullback entry, since the name is likely already partially owned. Falsifier: any guide cut or cloud-customer digestion signal from MSFT/AMZN.
• Long VRT versus short DELL as a pair trade for the next 1-2 quarters: prefer the power/cooling tollbooth over the server OEM with lower margin quality and more working-capital drag. Exit if DELL converts backlog to FCF faster than expected.
• Small tactical long LITE on a 3-6 month view for the highest beta to optical upgrades and 1.6T ramp, but size modestly because supply relief could cap the rerating. Risk/reward is attractive only if execution remains tight.
• Avoid chasing NVDA here; the marginal upside from this article accrues to non-chip bottlenecks, so the better expression is basket rotation into CLS/VRT/LITE rather than adding to the crowded compute trade.
• Set a watch item on the next AMZN/MSFT capex commentary: any sign of capex normalization should trigger profit-taking across DELL/CLS/VRT, while a second straight acceleration would extend the trade by another 1-2 quarters.