The race to miniaturize transistors has intensified as artificial intelligence demands increasingly powerful computing capabilities. Transistors are microscopic electronic switches etched into silicon chips that control the flow of electrical current billions of times per second, and the more transistors that fit onto a single chip, the greater its processing power and speed.
Researchers have achieved remarkable breakthroughs in transistor size reduction. Scientists recently demonstrated single-atom switches that control electron flow, and in a June study, researchers created nanoribbon transistors with channel widths of just 25 nanometers using two-dimensional semiconductors made from materials like tungsten disulfide. However, challenges remain beyond achieving smaller sizes. Manufacturing billions of transistors reliably at scale while keeping costs reasonable presents significant hurdles, and engineers must also address energy efficiency concerns as adding more transistors increases power consumption proportionally.
While traditional silicon transistors are expected to continue shrinking over the next decade, though more gradually than in the past, the industry may eventually need to transition to atomically thin materials for substantial further reductions. Additionally, engineers are exploring three-dimensional chip designs that stack transistors vertically to increase computational power without requiring proportionally smaller individual components. The ultimate goal remains balancing smaller size, increased speed, affordability, and energy efficiency simultaneously.
Originally reported by Livescience →
About the Author
