A large costs wave to tame. Underlying demand is saturated and now contracting under price pressure. Both core categories are set for their steepest drop in over a decade in 2026 – PC shipments down roughly 11% and smartphones down 14% – as rising prices push buyers to defer, lengthening PC lifetimes by 15–20%. The contraction is highly uneven: the entry level is collapsing (sub-USD500 PCs and sub-USD200 handsets losing ~20% of volume), while premium demand above USD800 is largely insulated. The mix is therefore shifting hard toward premium, where pricing power and ecosystem lock-in let vendors pass costs through, and toward the one growth pocket – foldables, up ~20% and the only segment where units and premium pricing rise together. The swing factor is B2C and B2B AI-hardware demand – an AI-PC and AI-smartphone refresh could reset upgrade cycles, but in an inflationary, higher-rate environment that catalyst is real yet delayed, its timing the single biggest variable for the topline.
The “RAMageddon” is threatening low-end hardware models as profitability erodes dramatically. The profitability outlook in 2026/2027 is dominated by one input: the heavy price of memory semiconductors. The AI data-center boom is siphoning DRAM, NAND and mobile memory capacity out of consumer devices, and the cost transfer is severe: memory is set to peak near 23% of PC bill-of-materials, up from 16% in 2025, with one major maker citing memory at ~35% of laptop material cost, up from 15–18% a quarter earlier, and notebook build costs potentially rising ~40%. This removes vendors' ability to absorb costs, forcing device prices up ~17% for PCs and ~13% for smartphones.
The margin consequence bifurcates the field: Premium and scale players with pass-through pricing and supply leverage protect margins and gain share, while budget-focused assemblers face outright margin destruction. Contract manufacturers – already operating at ~2–4% margins – are most exposed, since higher component prices inflate revenue without lifting assembly fees. The structural danger is that a permanent step-up in the cost base (memory plus tariffs) converts a cyclical squeeze into a solvency and liquidity problem for weaker, thinly-capitalized players, accelerating consolidation toward those who can fund working capital through the shock.
Ecosystem strength challenged by changing market conditions. The sector's persistent structural weakness is where it sits in the value chain: it captures little of the value it assembles. Profit is hyper-concentrated in a single premium ecosystem player, while component suppliers take the rents and the rest of the field fights over commoditized volume. This is a labor-intensive, easily-replicated activity with fragmented competition and thin margins – precisely the conditions under which a cost shock triggers shake-out rather than shared pain. Expect intensifying market concentration as margin and liquidity stress force weaker brands and subscale assemblers out, and as scale, procurement leverage and pricing power become decisive. The competitive advantage accrues to those who own an ecosystem (recurring services, accessories, software) or command the manufacturing scale to negotiate component allocation – everyone else is a price-taker on both sides. One structural response worth watching: OEMs exploring new topline channels beyond the box – device leasing, device-as-a-service and paid-in software/services – to add recurring revenue and defend margins as unit economics deteriorate.
A supply-chain transformation to monitor: de-risking China strategy to gain traction. The physical supply base is both a strength and the sector's sharpest structural risk. Its strength is a dense, flexible Asian supplier ecosystem capable of rapid, high-quality, high-volume execution, and activities that are relatively easy to automate and relocate when conditions change. The weakness is concentration: assembly was historically ~90% China-based, and finished devices are physically traded and directly tariff-exposed – a step-change from the tariff-immune software layer. The result is a forced, costly diversification: production is shifting to India (which assembled ~25% of the leading handset's output in 2025, with US-bound volumes targeted to move there by end-2026 to dodge tariffs), Vietnam and Mexico. But relocation is a multi-year capex drag, not a clean fix — non-China value-add remains only ~15–20%, with most components still imported, so the supply chain is de-risked geographically while staying dependent on the same upstream sources. Offshoring/nearshoring investment and reducing China’s footprint will be a defining capital-allocation theme, and export controls add a second fracture, cutting some brands off from Western operating systems and chips.
AI development still offers some promising signals. The same force squeezing the sector is also its efficiency lever. On the demand side, on-device AI is the primary hope for reviving replacement cycles, but adoption is being throttled by the very memory inflation that AI data centers are causing – cannibalization of global memory chip supply by AI infrastructure is delaying mass AI-PC penetration. On the supply side, AI integration into manufacturing, demand forecasting and inventory management is a genuine cost-mitigation route, helping offset input inflation and reduce the destocking write-downs that have historically amplified the sector's cyclicality. Slow-to-moderate capex intensity – a function of automation level – gives makers some flexibility to invest selectively without the balance-sheet burden of the capital-heavy silicon and network-equipment layers.
