Duke engineers show how a common device architecture used to test 2D transistors overstates their performance prospects in real-world devices.
Lab architecture used to test 2D semiconductors artificially boosts performance metrics, making it harder to assess whether these materials can truly replace silicon.