The debate over Micron Technology‘s (NASDAQ: MU) future gets incorrectly framed all of the time. Bulls point to artificial intelligence (AI) model scale and data center buildouts. Meanwhile, bears remain skeptical of the cyclical nature of memory chips and warn of inventory gluts.
I think both camps are indexing too heavily on yesterday’s demand curve. The question that smart investors should be asking is whether the next leg of memory demand is structural or speculative. The answer to that will shape where Micron stock trades by next decade.
Will AI create the world’s first trillionaire? Our team just released a report on the one little-known company, called an “Indispensable Monopoly” providing the critical technology Nvidia and Intel both need. Continue »
When it comes to memory chips, training large language models is all anyone seems to be discussing. But inference is where demand for memory truly compounds. Each time a user engages with a deployed AI system, the underlying hardware has to retrieve, process, and return data at low latency. Not only is this workflow incredibly memory-intensive, it scales with usage, not just expanding model sizes.
As AI applications shift from pilots to production environments, inference multiplies by orders of magnitude. Unlike model training — which is episodic — inference phases are ongoing and constantly expanding. Micron’s HBM3E and LPDDR5X products are positioned for this transition, not because AI models are getting bigger, but because deployed AI is nonstop.
The most visible narrative currently hanging over the AI ecosystem is data center construction. What rarely gets priced in are applications at the edge. Autonomous vehicles, smart manufacturing floors, and surgical robotics all require on-device memory capable of processing compressed AI models locally. This is completely different from high-bandwidth memory (HBM), as it runs on LPDDR and embedded NAND.
If edge AI adoption lands anywhere near the trillion-dollar projections for automotive OEMs and industrial equipment makers, Micron quietly gains a second, more lucrative demand vector that’s decoupled from the volatility of hyperscaler capex cycles. This decoupling is almost certainly undervalued right now.
Stock market bubbles are characterized by demand that only exists on paper. Micron’s revenue and profit growth are being driven by concrete purchase orders, not by projections inside of spreadsheets.
