Micron's CEO predicts self-driving cars will consume up to 300GB of RAM.

Looking at Apple's example, when the demand for running local AI models increases, Macs with 256GB or 512GB of integrated storage are always in short supply and their prices increase.

Micron's revenue "takes off" thanks to the AI ​​boom.

Micron has just released its Q2 2026 financial report with figures that have stunned the tech world. The company's revenue reached $23.86 billion, a phenomenal 200% increase compared to the same period last year ($8.03 billion). The main driver of this boom is none other than the seemingly insatiable demand for high-bandwidth memory (HBM) chips from AI service giants.

To meet this demand, Micron is aggressively expanding its "megafactories" in Japan, Singapore, and New York. However, according to CEO Sanjay Mehrotra, a new market that consumes even more RAM than servers is: self-driving cars.

Level 4 autonomous "monster": Requires 300GB of RAM to process data.

Currently, most modern cars only need about 16GB of RAM for basic functions. However, when it comes to Level 4 autonomous driving (L4) — a level where cars can handle complex situations like overtaking and crossing intersections without human intervention — the story will be completely different.

The risk of a new memory chip crisis.

Looking at the lesson from Apple, when the demand for running local AI models increased, Macs with 256GB or 512GB of unified memory were always in short supply and prices skyrocketed. If automakers start producing millions of L4 vehicles, the world could face a memory chip shortage worse than ever before.

Although Micron's super-factories are expected to become operational between 2028 and 2029, the pace of AI development in cars may be faster than anticipated. These "mobile supercomputers" promise to completely transform transportation, but also pose a significant challenge to the global component supply chain.