Key Points
-
For nearly three years, AI hyperscalers raced to procure as many GPUs from Nvidia as possible.
-
While GPUs remain an important pillar supporting AI development, larger models now demand next-generation memory and storage.
-
Micron Technology has emerged as a key supplier of high-bandwidth memory (HBM) for big tech.
Nvidia, primarily recognized as a graphics chip designer, has become essential in AI development, with significant demand for its products propelling it to become the most valuable company globally. From 2023 to 2025, Nvidia’s chipsets have underpinned generative AI operations, yet as of April 2, its stock has declined by 5.2% this year.
Micron Technology, on the other hand, has seen its shares surge by 27% year-to-date and nearly 309% over the past year, driven by the increasing demand for high-bandwidth memory to support AI models. The big five hyperscalers—Microsoft, Amazon, Alphabet, Meta Platforms, and Oracle—are projected to invest $700 billion in AI infrastructure this year, with Micron announcing it is sold out of its 2026 HBM inventory.
