Institute for Communication Technologies and Embedded Systems

SPRIND: Neureka - Computing-in-Memory Edge-AI System

To implement our visions for the future - from autonomous driving and computer-aided drug development to intelligent control of countless renewable energy sources - fundamentally new computing concepts will have to be found. After all, if global energy consumption for computing and communications increases at the same rate as it is now, it will take up the entire global capacity for energy production as early as 2040 with current hardware. Moreover, new application scenarios also place new demands on computing hardware that can only be implemented inadequately with current computer architectures. For example, to be usefully mobile in our everyday lives, so-called edge devices must only consume a minimum of energy.

ICE participates in the SPRIN-D Challenge “New Computing Concepts” in a competitive landscape with 14 other teams. Within this project, we will research a disruptive computing concept by a system-level compiler-architecture co-optimization using emerging computing-in-memory devices for AI at the edge to improve latency and power efficiency beyond the capabilities of von Neumann computing.

Further information about the challenge can be found at https://www.sprind.org/de/challenges/newcomputing

 

About SPRIND Challenges

SPRIND Challenges are innovation competitions that aim to generate solutions to the grand societal and technological challenges of our time. They create the vision of a better future and gather the scientists, innovators and entrepreneurs who can make this vision a reality. That's why Challenge Teams are funded quickly and without red tape, and immediately launch into a multi-stage competition. At the end of each stage, the teams' work is evaluated and only the best remain in the Challenge and receive further financial support to develop their idea.