TSMC and chip design software companies use artificial intelligence to help chips reduce energy consumption.

Computing chips that power artificial intelligence require a large amount of electricity. This Wednesday, the world's largest chip manufacturers showcased a new strategy to improve chip efficiency: using AI-driven software to design them. At a conference held in Silicon Valley, TSMC, a contract manufacturer for Nvidia, demonstrated a series of methods aimed at increasing the energy efficiency of AI computing chips by about 10 times. For example, Nvidia's flagship AI server currently consumes up to 1,200 watts of power when performing high-demand tasks, which, if run continuously, is equivalent to the electricity usage of 1,000 American households. TSMC aims to achieve the benefits from a new generation of chip designs where multiple "cores" are packaged together using different techniques to form a computing package. However, in order to utilize these techniques, companies designing chips are increasingly relying on AI software provided by suppliers such as Cadence Design Systems and Synopsys, both of which unveiled new products developed in close collaboration with TSMC on Wednesday.