TrendForce’s latest research reveals that rising demand for AI servers, general-purpose servers, and edge AI devices is prompting foundries to dedicate an increasing share of wafer capacity to AI-related products. This shift has significantly altered the supply-and-demand balance for mature process nodes.
TrendForce's latest research on power architectures for AI servers highlights that NVIDIA has been developing its proprietary 800V HVDC Power Rack. The platform is expected to be ready for customer shipments in 3Q26 and will be offered as an optional configuration—rather than a standard feature—for customers deploying the Vera Rubin platform.
TrendForce’s latest research reveals that structural tightening in mature-node DRAM supply is forcing consumer DRAM buyers to adopt legacy memory products to secure larger supply allocations. This has triggered a new wave of demand for procuring older-generation DRAM components, driving continued price momentum across legacy products such as DDR2 and DDR3.
The rapid growth of demand for AI semiconductors is driving the evolution of advanced packaging technologies, with Fan-Out Panel-Level Packaging (FOPLP) emerging as a new battleground across the industry. TrendForce reports that TSMC is currently focusing on its Chip-on-Panel-on-Substrate (CoPoS) packaging architecture and has standardized on a 310 × 310 mm panel format.
According to TrendForce’s latest MLCC industry research, the ongoing AI arms race among global CSPs is accelerating the adoption of in-house ASIC accelerators, which increasingly rely on small-form-factor, high-capacitance, high-temperature-resistant MLCCs. As such, demand is rapidly concentrating on a limited number of premium specifications. However, with supplier capacity expansion lagging behind demand growth, the risk of structural shortages in the second half of 2026 is becoming increasingly difficult to ignore.