电子通胀
Search documents
国金证券:谁的产能被AI挤占?从电子布出发,看好电子通胀强周期
智通财经网· 2026-02-24 04:00
Core Viewpoint - The report from Guojin Securities highlights that the rapid rise in prices within traditional industries is driven by the high returns from AI, which is reallocating resources and creating supply gaps. The industry is currently in a stocking phase, further catalyzing price increases, particularly in electronic materials [1]. Group 1: Impact of AI on Traditional Industries - AI is attracting capital due to its high return potential, leading to a significant reallocation of resources from traditional sectors, creating supply shortages [2]. - The competition in the industry is intensifying as companies upgrade their capacities to maintain technological advantages, resulting in a rapid evolution of the sector [2]. - Traditional capacities being directly squeezed by AI include storage, electronic fabrics, optical fibers, CTE electronic fabrics, CCL, CPUs, copper foils, packaging, passive components, power supplies, PCBs, and ATE [2]. Group 2: Specific Industry Insights - In storage, HBM is consuming wafer capacity at a much higher rate than traditional DRAM [3]. - For electronic fabrics, low-dk/low-cte/Q fabrics are taking over the capacity of 7628/thin/ultra-thin fabrics due to longer ordering cycles [3]. - AI data centers are consuming optical fiber capacity, particularly G.652D, due to a shortage of optical preform rods [4]. Group 3: Price Increases and Market Dynamics - The price increases in traditional sectors are occurring rapidly, with examples including DDR4 and LPDDR4 prices rising since May 2025, and ordinary electronic fabrics experiencing multiple price hikes throughout 2025 and early 2026 [10][11]. - The stocking behavior in the electronic materials supply chain is accelerating price increases, as companies anticipate further price hikes and react by increasing their inventory [11].