Analysts at CITIC Construction Investment published a morning note from which 36Kr pulled the key takeaway: the electronics sector is growing, and semiconductor localization could accelerate. The note's logic is simple. Memory is in an upswing right now, and that upswing is pulling capacity expansion across the whole industry. On top of that come investments in advanced logic nodes, contract manufacturing, and advanced packaging. Together, all of this is driving up global spending on chip manufacturing — and within that wave, two segments are growing: equipment and materials.
Here's the mechanism the team laid out. On the equipment side, growth is happening along two tracks: new fab capacity is being built, and memory makers are upgrading their process nodes in parallel. Each of those steps means new production lines — lines where Chinese equipment gets validated and gets a shot at volume deployment. The materials picture is different. Since the start of 2026, supplies of mature-node materials from abroad have been shrinking, while advanced lines inside China are expanding. Fab utilization on logic and specialty processes is holding at its peak, plus expansion is accelerating at memory makers and packaging houses — and materials demand is moving into a growth phase.
Look at what's actually being said here. This isn't a product story — it's a research seller's view of how demand for fab "hardware" works. The line about shrinking foreign supplies of mature-node materials is the most interesting part. If it's true, Chinese materials and equipment makers aren't getting a market-driven buyer — they're getting a captive one: the fabs simply have nothing else to buy. Our read: it's this mechanism, not some abstract "patriotic procurement," that's keeping utilization at its peak.
What does this actually change? Equipment and materials are the most inertial part of the industry: validation cycles are long, but once a supplier is in production, it's hard to kick them out. That's why the 2026 window matters more than this year's growth itself. If a Chinese line passes validation at an expanding fab today, it takes share in the next cycle tomorrow. For the market, that means orders shifting away from Western suppliers; for the wallet of the end-chip buyer, it means memory prices are unlikely to fall quickly — capacity is expanding because of high demand, not against it.
Who wins: Chinese equipment and materials makers, and the fabs that need to plug gaps in their supply chains. Who loses: foreign suppliers in mature nodes, if the restrictions don't ease. But that's our interpretation, not the note's conclusion — CITIC Construction Investment talks about an "opportunity to accelerate," not a done deal.
The weak spot here is the total absence of numbers. No capex volumes, no growth percentages, no data on specific fabs or memory plants, not a single company named. The claim that "utilization is at its peak" is an analyst judgment with no supporting statistics attached — readers can't verify it themselves. And there's no confirmation from the manufacturers themselves anywhere in the note.
Could capacity expansion driven by high memory demand turn into a glut a year or two from now — or will demand absorb everything that gets built?
Can domestic suppliers make up for the drop in foreign mature-node supplies when the question is quality and yields, not just volume?
Should the "validation window" count as a real advantage if customers are buying the equipment not by choice, but because there's no alternative?
Forecast: within two quarters, we expect at least one public announcement of volume deployment of Chinese equipment at an expanding memory fab — a verifiable sign that the note's thesis is holding up in practice, not just on paper.
Disclaimer: a single source — a CITIC Construction Investment analyst note as relayed by 36Kr; the piece contains no official statistics on fab utilization or capex.
Source: www.36kr.com · post in Telegram