Why AI Could Cause a Chip Crunch

I’ve just pre-ordered what I want from the January 6-9 2026 Las Vegas Consumer Electronics Show. My first choice (from two articles that list their favorites) is the Pebble Index 01. As a ring that records, it will let me note ideas for blogs, classes, books before I forget them. (My phone and watch have not been great for verbal notes):

Luckily, DRAM will not be an issue.

An (AI) Supply Crunch for Chips

Producers of our everyday electronics could have a supply problem. DRAM (dynamic random-access memory) users, the memory in our smartphones, laptops, and assorted other electronics, could be nudged aside by the chips needed for AI data centers. Fueling AI, HBM (high-bandwidth memory) chips need three to four as many silicon wafers as the standard DRAM using device. As a result, until manufacturers boost capacity, maybe in two years, consumers could face higher prices and delayed and diminished shipments. We should note that developers could shrink the problem with workarounds that Wired said were possible.

Classic supply and demand, equilibrium price is way up for DRAM users like laptop makers:

Our Bottom Line: Externalities

Because of an AI Data Center in Oregon (or elsewhere), my smart TV could cost me more or not even arrive. Defined as the impact of a decision or contract on a third uninvolved party, an externality can be positive or negative. The classic example for a positive externality is when vaccinated individuals keep unprotected individuals healthy. On the negative side, we have pollution that can sicken people that have no connection to the factory.

Correspondingly, with AI creating a DRAM shortage, you and I could pay more for a new self-parking electric vehicle or endure a longer wait for our new smart fridge.

My sources and more: The Economist and Wired are perfect for those of us that need to learn more about HBM and DRAM created shortages. Then for 2026 CES favorites, the NY Times and Wired had some possibilities.