Back
OCCAM
The recent technological and architectural innovations have shifted the main bottleneck of current and future processors from computing to communications. The OCCAM proposal hence focuses on the grand challenge of removing the persistent communication bottleneck that hinders the scaling of computing systems at the end of Moore’s Era. This problem is stalling the progress of well-established and emerging applications in the entire edge-cloud continuum, from on-device AI to Large Language Models, from commercial applications to scientific computing. The vision is hence the realization of massive heterogeneous computer architectures with native support for distributed collective all-to-all operations that, by merging communications and computing, bypass the interconnect bottleneck. The targeted breakthrough is the experimental validation of the concept of over-the-chip wave computing with mm-wave reconfigurable intelligent surfaces, as well as its architectural significance within next-generation computing systems.
Field of action:
CMOS and more
- Faculty 6 – Electrical Engineering and Information Technology
Address:
Lehrstuhl für Höchstfrequenzelektronik (HFE)
Contact:
Univ.-Prof. Dr. Renato Negra
Renato.negra@hfe.rwth-aachen.de
Homepage:
http://www.hfe.rwth-aachen.de
Status:
finished