ChipsThe story, in brief
IBM has unveiled chip technology that could help extend Moore’s Law another decade
100 billion transistors on a fingernail. IBM just bought Moore's Law another decade—and that changes everything for AI compute.
The KeyNews take
Why it matters
IBM's sub-1nm chip breakthrough directly impacts AI infrastructure economics. Denser, faster, more efficient transistors mean lower compute costs and higher performance ceilings for training and inference—critical for companies competing on model scale and deployment.
The key facts
5 to know100 billion transistors achieved at sub-1nm density
2x density improvement vs. IBM's 2021 state-of-the-art
Prototype stage (not production)
Implications for energy efficiency and compute performance in AI workloads
Extends Moore's Law trajectory by ~10 years per headline claim
Go to the source
MIT Technology Reviewtechnologyreview.com
Publisher excerpt: IBM has built a new prototype chip with around 100 billion transistors on an area the size of a fingernail, which is twice the density of the company’s previous state-of-the-art technology announced in 2021. The design could pave the way for faster and more energy efficient computers for years to…