MaxLinear at Citi’s 2026 Global TMT Conference: data center growth builds

MaxLinear said its optical data-center revenue guidance midpoint has risen to $220 million after three consecutive increases, while infrastructure—now more than 50% of revenue—is expected to grow 150%-170% in 2026. Management targets optical market share of 20%-25% within two to three years from roughly 5%-7% currently, supported by 1.6T and 3.2T product ramps, firm pricing, and higher system content. Gross margin is projected to approach 60% in 2026 versus 57.48% trailing-12-month margin, although wafer-price inflation and supply execution remain key risks.

Analysis

MXL’s upside now rests less on AI-optics demand—which is broadly understood—and more on converting design qualifications into a durable 1.6T share step-up without sacrificing gross margin. At its current valuation after a parabolic run, the market is implicitly underwriting management’s multi-year share ambition; any evidence that 1.6T qualification converts more slowly than expected, or that module vendors dual-source aggressively, creates disproportionate multiple risk. The next 1-3 months are therefore about order quality, backlog conversion and foundry economics rather than another headline guidance raise.

Samsung diversification is strategically positive only if yields, reliability and customer qualification match incumbent-node production. A second source can improve MXL’s allocation position versus optical peers dependent on TSMC, but prepayments, inventory buffers and wafer premiums could absorb much of the advertised mix-driven margin expansion before revenue scales. Watch days inventory, purchase commitments, gross-margin guide and any gap between bookings and shipments; a margin guide below 60% despite strong sales would signal that supply access is being bought rather than monetized.

The less obvious beneficiary is AMD: Panther provides an ecosystem feature that reduces a platform-level disadvantage versus Intel’s integrated compression/offload stack, potentially supporting AMD server attach and enterprise-storage design wins. Conversely, Intel’s risk is narrow and likely immaterial financially; the more relevant competitive read-through is whether MXL’s claimed power advantage forces incumbents to defend share through pricing, which would make the industry’s current ASP assumptions too optimistic. CPO/NPO remains a 6-18 month optionality story, not a near-term earnings driver, and architecture changes can redistribute content even if total content rises.

Contrarian view: MXL is becoming a crowded pure-play expression of AI networking scarcity. The asymmetric positive surprise is a named hyperscaler or top-tier module qualification plus sustained gross-margin expansion; the more probable near-term surprise is normalization from an unsustainably rapid upgrade cadence. Treat management’s long-range content and share targets as scenario inputs, not base-case estimates, until independent shipment and customer-design-win evidence confirms them.

Key Decisions for Investors

• Maintain MXL as a tactical long only on pullbacks of 15-20% or after the next earnings report validates both optical revenue conversion and a >=60% gross-margin outlook. Target 20-30% upside over 6-12 months if 1.6T ramps on schedule; exit or hedge if gross-margin guidance falls below 58% or 2027 optical commentary shifts from orders to forecasts.
• For existing MXL exposure, buy 3-6 month downside protection via put spreads rather than add spot after the sharp rerating. A qualification/yield delay can compress the stock materially even with secular AI-optics demand intact; finance selectively with out-of-the-money calls only if implied volatility is not already pricing a post-earnings move above 15-20%.
• Use a relative-value basket: long MXL / short a diversified semiconductor ETF such as SOXX for investors seeking company-specific 1.6T share-gain exposure while reducing broad AI-capex beta. Reassess after the next quarter on evidence of incremental customer concentration, Samsung yield performance, and bookings coverage.
• Watch AMD for Panther-related reference-design conversions, but do not establish a standalone trade on this linkage. Upgrade the thesis only if AMD or a major storage OEM discloses accelerator attach rates or enterprise platform wins; otherwise the revenue contribution remains too small to move AMD estimates.
• Monitor TSM capacity/pricing and Samsung advanced-node yield disclosures as leading indicators. An easing of optical component constraints would reduce MXL’s pricing umbrella, while a TSM allocation shock could make Samsung-backed supply strategically valuable but still pressure MXL working capital.