QM/EM for devices and systems
At quest we are interested in state-of-the-art (nano)electronic devices and systems, which nowadays operate at very high frequencies and are often also extremely miniaturised. Therefore, their behaviour is dominated by various electromagnetic (EM) and quantum mechanical (QM) phenomena. Our general goal is to model this QM/EM behaviour of novel (nano)devices and systems, in order to better understand the underlying physics and to support their design.
The research at quest can be roughly broken down into six (often overlapping) topics, as indicated below.
QM and hybrid QM/EM modelling of nanodevices
Owing to their extremely small size and/or to the presence of 1-D and 2-D materials such as carbon nanotubes, graphene, transition metal dichalcogenides, etc., the transport of charge carriers in nanoelectronic devices requires careful QM modelling. Moreover, in many cases, a QM/EM hybrid approach, where all pertinent EM effects are accounted for as well, is needed. We develop new QM and QM/EM modelling strategies for nanoelectronic devices and applications.
EM modelling of interconnects, packages, and devices
A major challenge in state-of-the-art electronic design lies in getting high-frequency signals of the die, through the package, and onto the printed circuit board. Modelling of such links is of critical importance and only full-wave approaches suffice, as these account for skin-effect, proximity-effect, slow-wave effect, etc. We construct full-wave solvers which handle the complex geometries, materials and EM phenomena encountered in packages and interconnect structures.
EM modelling for scattering and antenna systems
EM modelling of scattering and antennas has been a long-standing research topic. Nonetheless, with the advent of 5G and 6G applications, huge challenges in this domain are still to be tackled. This is owing to the complex geometries and the high operating frequencies. In collaboration with UGent’s EM Group, who are experts in antenna design, we develop novel computational EM algorithms for large-scale and intricate antenna and scattering problems.
Electromagnetic Combatibility, Signal and Power Integrity
The Electromagnetic Compatibility (EMC) behaviour of a device requires careful monitoring during all stages of a design, not only because of the stringent EMC regulations, but also since it influences the functional behaviour of an electronic product. Indeed, EMC-behaviour is strongly intertwined with the Signal and Power Integrity (SI/PI) of a device. At quest, we develop EMC/SI/PI-aware modelling and design strategies.
Uncertainty quantification of (nano)devices and systems
All electronic devices and systems are prone to manufacturing tolerances, causing variability of their geometrical and material parameters. This leads to uncertainty of the manufactured device’s EM behaviour. This effect is exacerbated by miniaturisation of state-of-the-art devices, where manufacturing variability is observed down to the QM level. We develop uncertainty quantification (UQ) techniques for (nano)devices and systems.
Optimization and design of (nano)devices and systems
All electronic products are the result of a careful design and optimization process. Given the high complexity of the modern QM/EM devices and systems, a successful design critically depends on the availability of accurate and efficient computer-aided design tools. In collaboration with UGent’s SUMO Lab, we leverage machine learning (ML) techniques to expedite optimization and design processes of (nano)devices and systems.
Developing high-performance models stands or falls with thorough validation and testing. Therefore, if feasible, we strive to incorporate experimental validation into our model development beside extensive simulation tests. To this end, and also for our work on EMC/SI/PI-aware design, we share a state-of-the-art EM lab with UGent’s EM Group.
The lab is organized around two anechoic chambers, one measuring 8.1m x 3.9m x 3.6m, while the other has dimensions of 6.4m x 4.0m x 3.7m, each equipped with a positioning system. Supporting test equipment includes a 4-port VNA up to 67 GHz, mmWave range extenders from 55 GHz to 170 GHz, a signal analyzer extendable up to 60 GHz, a 60 GHz arbitrary waveform generator, real-time oscilloscopes, …
For debugging and performing validation runs, quest owns high-performance multiprocessor computation nodes. Larger simulations and evaluations are relegated to Ghent University’s supercomputer or to the Flemish Tier 1 HPC.