Elon Musk’s Colossal Chip Demand: Ambitious Vision or Miscommunication?

Elon Musk continues to fuel discussions about the sheer number of chips his companies require for AI training. The conversation has now reached specific numbers, which are as bold as some of the billionaire’s other ambitions. Musk stated that Tesla needs between 100 to 200 billion chips annually.

This isn’t about all the chips Tesla needs for its vehicles, but specifically about AI accelerators. “I have immense respect for TSMC and Samsung. We work with both TSMC and Samsung in Tesla and SpaceX. They are excellent companies, and we want them to produce our chips as quickly as possible and scale them to the maximum volume that they are comfortable with. But apparently, that’s not enough,” Musk said.

When I asked how long it would take to build a new chip manufacturing plant, they replied that it would take five years to launch production. To me, five years is an eternity. My timelines are one, two years. I can’t even imagine a three-year period. That won’t be enough. If they reconsider and say, yes, they will work faster and will supply us with 100, 200 billion AI chips per year at the times we need them, that would be great.

To understand how incredible these numbers are, it is enough to know that around 1.5 to 2 trillion chips are produced globally every year, but this includes all types of chips, primarily simple controllers. In the case of AI accelerators, even Nvidia, which dominates a significant part of the market, supplies only a few million units annually. This indicates a difference of approximately 10,000 times.

There is a possibility that Musk was not thinking about quantity but about monetary value, referring instead to $100-200 billion that Tesla might spend on chips annually. This would completely change the conversation, although these are still gigantic sums. Regardless, Musk has already stated that Tesla will need to build its own gigafactory for chip production, and certain steps have already been made.

New Developments and Market Dynamics

As of late 2025, efforts to accelerate AI chip production have intensified globally. Nvidia’s latest advancements in AI accelerators boast improved efficiency, yet they still face the challenge of matching the massive demand suggested by Musk. Meanwhile, TSMC and Samsung are exploring expansions in their production lines, with Samsung recently announcing a potential new plant dedicated to semiconductor advancements.

Globally, the trend toward increasing semiconductor production continues with giants like Intel and Taiwan’s TSMC investing significantly in infrastructure that promises future-ready manufacturing capabilities. Yet, the massive numbers Musk mentions remain an outlier, hinting perhaps more at the fiscal capacity Tesla earmarks for these components.