-
-efficient accelerator architectures by designing digital and mixed-signal blocks for SSM inference in edge-AI systems. The successful candidate will explore novel SSM architectures that support sequence
-
-design SSM-inspired spiking neural network (SNN) cores and integrate them with a low-power RISC-V processor. The PhD candidate will develop spike-based computing blocks and explore hybrid analog/digital
-
to develop circuits that convert analog neural signals into digital or event-based outputs (e.g., spike-compatible signals) suitable for a neuromorphic backend. Note: This position does not involve designing
-
include: CMOS-based neuron and synapse circuit design Low-power digital architecture for SNN processing On-chip learning mechanisms Integration with sensor interfaces for biomedical signal processing What
-
allocation, start-up/shutdown sequences, and operational planning of PtX systems in response to dynamic market and process conditions. System Integration and Digital Twin Development: Combine experimental
-
, modularisation and platform design. Experience with Digital Advanced Product Modelling using CAD design, simulations, and mathematics. A strong motivation for collaborative projects within academia and industry
-
nearing completion) in Electrical Engineering, electronics engineering, power electronics, or a related field. Basic knowledge of Basic knowledge of power converter topologies. Basic knowledge of digital
-
, power electronics, or a related field. Basic knowledge of power converter topologies. Basic knowledge of digital control of power converters Experience or coursework in device simulation tools (e.g
-
-out key analog and mixed-signal building blocks (e.g., low-noise amplifiers, voltage/current sources, multiplexers, digital logic) for operation at cryogenic temperatures. Develop measurement setups
-
for resilient high-mix low-volume manufacturing. The aim of this PhD project is to enable fast setup of robot manipulators to complete advanced manufacturing tasks by the use of digital models. This should be