PhD in digital accelerator design for computational neuroscience 100%
Publication date:
29 April 2025Workload:
100%- Place of work:Bern
University of Bern, Department of Physiology
01.06.2025 or by agreement
PhD contract for 4 years
Extended visits to Yale University are envisioned
We are seeking a highly motivated PhD candidate in accelerator design for computational neuroscience as part of a joint project between the Neuro-inspired Theory, Modeling and Applications (NeuroTMA) Lab led by Mihai A. Petrovici at the University of Bern and the Asynchronous VLSI and Architecture (AVLSI) Lab led by Rajit Manohar at Yale University.
The NeuroTMA Lab focuses on biologically inspired models of spatio-temporal processing in recurrent cortical networks and their relation to reinforcement learning, Bayesian computing, and neuromorphic implementations. The AVLSI Lab conducts research on semantics, design methodologies, and architectures for asynchronous systems with the goal of designing and implementing efficient and programmable computation structures.
The project will develop an accelerator for the simulation of cortical models of processing and learning.
Tasks
The NeuroTMA Lab has recently developed a novel model of cortical processing and
learning that describes the relevant biophysical processes evolving in continuous time through systems of differential equations. This model lends itself to a highly efficient hardware implementation capable of real-time on-line learning from various kinds of input data streams. Our project seeks to carry out such an implementation by building on a novel neuromorphic architecture co-developed by the AVLSI Lab.
You will contribute to the design, implementation, tape-out, and verification of a custom digital chip implementing a massively parallel, asynchronous differential equation solver. Furthermore, you will work on extending our existing theory of on-line learning to more complex single neuron dynamics, for example oscillators. To this end, you will collaborate with other researchers and faculty in an interdisciplinary setting and be mentored by an experienced postdoctoral fellow.
Requirements
A master's degree in electrical engineering, computer science, physics or a closely related field is required. You are expected to have demonstrated an excellent grasp on quantitative methods (dynamical systems theory, numerical methods, foundations of machine learning, etc.), evidenced by exemplary academic performance and/or outstanding project work. You will need to be able to communicate and collaborate effectively across discipline boundaries and have a well-developed attention to detail.
Ideal candidates have a strong enthusiasm for interdisciplinary research and are naturally curious, eager to continuously grow their knowledge and skillset. Experience in digital hardware development, from design to tape-out and proficiency with associated EDA and simulation tools are advantageous.
We offer
Application and contact
If you are interested, please send one .pdf document including your CV (max two pages), a statement of research interests (max one page), at least one single-author work (e.g., bachelor's/master's thesis), and the name and contact details of two references to E-Mail schreiben, E-Mail schreiben and E-Mail schreiben.
Review of applications will begin immediately and continue until the position is filled. Informal inquiries are welcome.
