You’ll never be stuck with a lack of memory. Google is not immune.
AI companies are competing for ever-increasing quantities of memory chips. problem? This industry is highly supply constrained. Costs are rising, products are tied up, and some companies, especially those in consumer electronics, are raising prices.
On the AI front, Google DeepMind CEO Demis Hassabis told CNBC that physical challenges are “limiting a lot of adoption.” Google sees “much more demand” for Gemini and other models than it can accommodate, he said.
“It also limits the research a little bit,” Hassabis says. “We need a lot of chips to be able to experiment with new ideas at a large enough scale to see if they actually work.”
Whether you work for Google, Meta, OpenAI, or another big tech company, researchers want chips, and memory is a critical component. Mark Zuckerberg said that AI researchers are asking for two things besides money: minimize the number of people reporting, and include as many chips as possible.
Hassabis said there are always “choke points” where there are capacity constraints.
“The whole supply chain is under strain,” Hassabis said. “We’re lucky because we have our own TPU, we have our own chip design.”
Google has long been building TPUs (Tensor Processing Units) for internal use. This also puts Nvidia at risk because the company also leases them to external customers through the cloud.
But even with access to its own TPUs, Google doesn’t have to navigate a competitive memory market. “It comes down to a few suppliers supplying some key components,” Hassabis said.
Three suppliers dominate memory chip production: Samsung, Micron, and SK Hynix. These companies are struggling to meet demand for chips from AI hyperscalers without losing longtime electronics customers.
It doesn’t help that AI companies primarily want different types of memory chips than PC manufacturers. Creators of large language models want HBM (High Bandwidth Memory) chips.
Don’t expect Google’s spending on AI infrastructure and chips to drop anytime soon. In its fourth-quarter earnings release, the company predicted capital investment in 2026 to be between $175 billion and $185 billion.
