Physical Computation: From Neural Morphology to Silicon
(and beyond) Substrates

Prof. Dr. Melika Payvand
Neuroinfomatics, University of Zurich, Switzerland

Understanding how physical form gives rise to function in the brain offers powerful clues for designing efficient computing systems. In the Emerging Intelligent Substrates Lab, we study how neural morphology, connectivity, and dynamics can be abstracted, emulated, or rediscovered in silicon.
I will first discuss how embedding biologically grounded constraints, such as dendritic integration, sparse local connectivity, and temporal computation, acts as an inductive bias
that improves generalization, power efficiency, and area utilization in neuromorphic hardware.
I will then consider the inverse question: when we impose functional and energetic constraints on artificial systems and allow structure to emerge, do we recover motifs reminiscent of neural morphology? Comparing such emergent forms to those in the brain may reveal why both biological and silicon systems converge toward similar organizational principles.
This bidirectional perspective highlights how physical form and computational function coevolve, guiding both our understanding of neural systems and the design of future computing substrates.

CRC 1461: Neurotronics
Colloquium: 28-October-2025_38, Tuesday, 10:00 am to 11:00 am (CET)

Link to the zoom meeting

Invited by Wilhelm Braun
Kiel University, Faculty of Engineering, Department of Electrical and Information Engineering

Download the announcement here.