As soon as possible

Ferroelectric Field-Effect Transistors (FeFETs) are considered as one of the enabling technologies for the next generation highly energy efficient computing systems. FeFETs are programmable, non-volatile silicon-based devices that can facilitate novel architectures able to efficiently perform complex, irregular, highly dynamic computational tasks. More precisely, our focus in on bringing processing capabilities close to the data sources, e.g., AI at the edge but not limited to. The goal is to minimize data access latency, bring the overall system energy efficiency to levels impossible with current technology and as a result enable new applications. In this PhD project you will work on the design and validation of a new generation of FeFET-based data processing architectures, algorithms and tools that can accelerate complex applications able to solve challenging problems from real-life. The work will be performed within a multidisciplinary team of PhD students working at different levels, starting from advanced ferroelectric materials and circuit design, up to heterogeneous computing systems.

MSc degree in the field of computer engineering, computer science, electrical engineering or similar with an affection for computationally demanding applications and machine learning.

Prof. dr. ir. Georgi Gaydadjiev (g.gaydadjiev@rug.nl)

Suppose optical signals are encoded as 2D datasets, i.e. as an image with information in intensity, wavelength and polarization. How do you then design metasurfaces to perform common image processing steps, nowadays done on a PC, optimally in the optical domain? Can you use metasurfaces for all-optical feature extraction and classification tasks? What are the limits to stacking metasurfaces for more complicated algorithms? You will perform electromagnetic design and optimization and demonstrate concepts in optical experiments.

Master’s degree in physics, photonics, nanoscience, or a related field, and with an affinity to both experiment and theory.

PHD Position: Metasurfaces for energy-efficient wave-based computing - AMOLF

as soon as possible

All-Optical switching (AOS) of magnetization was demonstrated as a radically new approach to write magnetic bits at an unprecedented ps timescale and with low switching energies. Recently we have exploited multishot helicity dependent AOS and realized opto-magnetic synapses in CoPt films: non-volatile magnetic memory elements continuously adaptable with laser pulses. Key challenges for this project are to develop material systems where AOS can lead to multi-state switching, can be scaled down to nanometer dimensions and can be integrated in a photonic network while still retaining the energy efficiency of AOS.

MSc degree in Experimental Physics or equivalent qualification

Prof. Theo Rasing (theo.rasing@ru.nl)

Can we make metasurfaces, which can process information encoded in structured optical beams ,programmable, trainable and dynamically reconfigurable? Can you switch between metasurface functions by incorporating suitable materials? If so, what are the limits in energy and speed? Can you train such active metasurfaces to realize all-optical neural network functions? We envision using novel materials as metasurface constituents, such as phase change materials that show strong and reversible changes in their linear and nonlinear optical response.

Master’s degree in physics, photonics, nanoscience, or a related field, and with an affinity to both experiment and theory.

PHD Position: Metasurfaces for energy-efficient wave-based computing - AMOLF

This project focuses on the demonstration of ferroelectric control over spin-orbit torques using two-dimensional materials for atomically-thin non-volatile data processing and storage.

The ideal candidate has:

• a master’s degree in physics, electrical engineering, materials science, or another related field.

• previous experience in optics, electronic transport, optoelectronics, magnetism, nanofabrication, or two-dimensional materials and heterostructures are a plus.

• some computer programming (e.g. Python) knowledge is appreciated.

• good spoken and written English proficiency is required.

Prof. dr. Marcos H. D. Guimaraes, m.h.guimaraes@rug.nl

as soon as possible

Although discovered much more recently than other types of Resistive RAM elements, ferroelectric memristors show the required performance for integration with CMOS. In the past couple of years, ferroelectric memristors with very promising properties have been demonstrated, as non-volatile resistive memory elements, including large OFF/ON ratios > 104 , reproducibility, high speed (< 10 ns), and potentially low operation energy (< 10 fJ). Most important, to minimize overall system power, the very high tunnelling electro resistance (TER) values (up to 1010 Ohms) of FTJs are ideal for integration with CMOS, resulting in low power operation.

In this project, crossbar arrays of memristive devices will be developed, in collaboration with TSST and Solmates, based on ferroelectric tunnel junctions (FTJs) and explore their integration with CMOS technologies will be explored. With this, we will demonstrate FTJs compatibility with CMOS electronics. FTJs are based on epitaxial strained oxide ferroelectrics and therefore require specific substrates and high processing temperatures. This hampers the integration of ferroelectric oxides with Si/CMOS. Lattice mismatch, thermal expansion mismatch and chemical stability of the substrate are the main difficulties to be overcome. Here, the challenge is to obtain epitaxial and ferroelectric tunnel barriers with the ferroelectric polarization out-of-plane. SrTiO3 has been regarded as a good template to deposit most of functional oxides, due to the low lattice mismatch with Si and chemical compatibility and films with substrate-like quality can be realized.

· You are an enthusiastic and highly motivated researcher.

· You have a MSc degree in Chemistry, Physics, Materials Science or equivalent, with excellent experimental skills.

· You have a creative mindset and excellent analytical and communication skills.

· You have a keen interest in Materials Science, Materials Chemistry and Physics.

· You like to work in an interdisciplinary and internationally oriented environment.

· You are fluent in English.

Prof. dr. ir. Gertjan Koster (g.koster@utwente.nl).