Hussam Amrouch, a professor at the Technical University of Munich, has developed a new AI architecture using special circuits known as ferroelectric field effect transistors (FeFETs). This architecture is twice as powerful and more energy efficient than previous approaches. The transistors on these new AI chips perform calculations and store data, making them faster and more efficient. The chips, developed in collaboration with Bosch and Fraunhofer IMPS, can perform 885 trillion operations per second per watt (TOPS/W), making them twice as powerful as other AI chips. The researchers designed the chip based on the architecture of the human brain, where neurons process signals and synapses store information. These FeFET transistors can store information even without power and allow for simultaneous storage and processing of data. The goal is to use these chips for deep learning algorithms, object recognition, and drone data processing. However, it will take a few more years before these chips are market-ready and meet the security requirements of industries like automotive. Interdisciplinary collaboration is crucial for the development of these chips, involving researchers from computer science, informatics, and electrical engineering.


>Source link>

>>Join our Facebook Group be part of community. <<

By hassani

Leave a Reply

Your email address will not be published. Required fields are marked *