The Ewald Marquardt Private Foundation has awarded its biennial Future Prize for the ninth time. The first prize of €10,000 was awarded to a team of developers from the Institute for Electrical and Optical Communications at the University of Stuttgart. Jakob Finkbeiner, Raphael Nägele and Dr. Markus Grözing have developed an analog computing circuit concept that greatly reduces power consumption and the chip area required for the basic arithmetic operations of multiplication and addition. In booming AI applications such as ChatGPT, these operations have to be performed billions of times to calculate so-called artificial neural networks in large data centers, consuming more and more energy in the process. On mobile devices such as smartphones, locally executed AI applications are draining the battery faster and faster.
This is where the new analog circuit concept from the University of Stuttgart comes in. In conventional digital arithmetic units, many hundreds of components, so-called field-effect transistors, are generally used for the simple multiplication of two numbers and are recharged for each calculation, which consumes a lot of energy during the calculation. In the new analog concept, multiplication is performed with just two field-effect transistors, in which a current flows for only a very brief moment. No additional component is even required for the addition. This results in a very high potential for increasing efficiency. Arithmetic units for AI applications on computer chips in data centers and smartphones could therefore become significantly smaller and more energy-efficient. Previous studies indicate that analogue multiplication with the new technology from the University of Stuttgart is around 100 times more energy-efficient than an already greatly reduced and optimized digital FP4 multiplication on the latest generation of Nvidia AI accelerators (GPU Blackwell GB200).
