2026-01-17 07:25:46

How many years will the hunger for computing power in the AI era last? The answer to this question might be glimpsed from the pace of global chip capacity expansion.

Recently, an important signal from the industry: major global semiconductor manufacturers announced new wafer fab projects, mainly focused on expanding high-performance storage chips—especially DRAM capacity. This will become the largest chip manufacturing facility in the region.

The problem is that the production start date has been pushed to 2030.

What does this mean? Currently, the demand for storage and computing chips for AI model training and data center expansion is extremely tight, but new capacity won't be available for another 5 years. How long this supply-demand gap will last and how the market will respond have become core issues that the industry chain upstream and downstream are contemplating. The lengthening of the chip cycle may redefine the landscape of hardware costs and the economics of computing power in the coming years.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

19 Likes

Reward
19
9
Repost
Share

Comment

0/400

MEVSupportGroup

· 01-19 05:06

Will it only be put into operation in 2030? That's a 5-year gap, and the computing power shortage will be at its worst during these years.

View OriginalReply0

ImpermanentSage

· 01-19 01:46

Will it only be put into production in 2030? Then during these 5 years, won't they starve to death?

View OriginalReply0

YieldWhisperer

· 01-18 07:03

actually the math doesn't check out here... 2030 for dram? that's assuming zero acceleration, zero alternative fab capacity coming online. i've seen this exact supply crunch narrative play out before—everyone panic buys, then suddenly 2028 hits and there's oversupply. classic boom-bust pattern.

Reply0

RealYieldWizard

· 01-17 07:46

Wait, it won't be operational until 2030? Doesn't that mean chip manufacturers and large model companies will be panicking? Artificial intelligence is really going to get stuck here; a five-year gap is a bit outrageous. It should have expanded production earlier; now rushing at the last minute is already too late. The DRAM shortage is real, but the 2030 timeline is a bit mystical—there must be other reasons. By then, the landscape will probably have changed; who knows what AI will develop into.

View OriginalReply0

MysteryBoxOpener

· 01-17 07:44

Capacity will only be released in 2030? Then the chip prices will skyrocket over these 5 years, and GPU miners will be making a fortune again, haha.

View OriginalReply0

BlindBoxVictim

· 01-17 07:42

Will it only be put into production in 2030? By then, AI will probably be dead from overcompetition... As for chips, it's basically a gamble on timing. Being five years late could mean a two-generation gap. Wait, isn't this just an excuse for a disguised price hike? Who profits the most from this supply shortage... I bet TSMC is laughing again. Really, hardware costs are probably going to skyrocket in the next few years, and retail investors will get cut again.

View OriginalReply0

SerNgmi

· 01-17 07:41

Will it only be put into production in 2030? Then the chip prices will have to skyrocket over these 5 years.

View OriginalReply0

SchrodingerGas

· 01-17 07:37

Production only in 2030? This gap will take 5 years, and DRAM prices are probably going to skyrocket... The equilibrium of the computing power economy has been pushed further back.

View OriginalReply0

MEVHunterNoLoss

· 01-17 07:35

Will it only be put into production in 2030? Who will fill this five-year gap? Chip stocks are bound to rise again.

View OriginalReply0