IBM scientists create tiny carbon nanotube transistors that could potentially replace silicon transistors in future computer processors
John Markoff / New York Times :
Context & Ripple Effects
This lands mid-arc for IBM's research division: months earlier it unveiled a functional 7nm test chip targeting a big power/performance jump over 10nm, and weeks earlier John Markoff reported engineers hunting new approaches as Moore's Law slows. The nanotube work is the answer to the question those pieces raised — what comes after silicon runs out of room.
It also prefigures the industry's own forecast: within a year the Semiconductor Industry Association projected transistors would stop shrinking around 2021, pushing gains toward vertical density and new materials rather than lateral scaling.
First-order effects
- IBM now holds both halves of a post-silicon hedge — an advanced silicon roadmap (7nm test silicon) and a candidate replacement technology in carbon nanotubes — giving its research arm a credible claim on whichever path wins.
Second-order effects
- Rival chipmakers and their equipment suppliers face pressure to fund parallel channel-material research, since a working nanotube transistor would devalue years of investment in ever-smaller silicon lithography if it scales to manufacturing.
Third-order effects
- If shrinking really stops on schedule, the industry's structure shifts from a single silicon scaling ladder to competing material-and-architecture bets (nanotubes, stacked 3D designs like the later 0.7nm nanostack process, vertical density), with leadership decided by whoever industrializes the successor first.
The trend: As Moore's Law's lateral shrink exhausts itself, chip innovation is migrating from smaller silicon features to new materials and 3D architectures — with IBM positioning across both.