IBM has unveiled what it describes as the world’s first sub-1 nanometre semiconductor technology, introducing a new 0.7nm (7-angstrom) chip architecture as the race to develop more powerful processors for artificial intelligence intensifies.
The announcement helped lift the Armonk, New York-based company’s shares by more than 6% in pre-market trading, although the stock remains down around 11% since the start of the year.
The breakthrough comes as chipmakers seek new ways to sustain the long-standing trend of increasing computing power while reducing component size. IBM’s latest technology is designed to strengthen its position in the semiconductor industry, where it competes with manufacturing giants such as TSMC and Intel.
According to IBM, the new 0.7nm technology can accommodate nearly 100 billion transistors on a surface roughly the size of a fingernail. This represents around twice the transistor density of the company’s 2nm chip, unveiled in 2021. The firm claims the design could deliver up to 50% greater performance or improve energy efficiency by as much as 70%.
A key feature of the innovation is a new transistor architecture known as “nanostack”. Rather than arranging transistors side by side on a flat surface, the technology stacks them vertically in three dimensions, allowing significantly more components to fit within the same physical space.
“Our new nanostack architecture is about far more than shrinking transistors,” said Jay Gambetta, Director of IBM Research. “It represents a fundamental rethink of chip design, enabling substantial gains in both computing power and energy efficiency.”
IBM believes commercial production could begin within the next five years. While the company has previously licensed semiconductor technologies to Samsung and Japan’s Rapidus, it has yet to reveal which manufacturing partner will produce the new chips.
