KAUST allows for "the chance to often tackle 'risky' research problems," Bagci said, which was a main motivator for him to join the University. As his research also focuses on the development of new simulation tools, Bagci was impressed by what the KAUST Supercomputing Core Lab could offer.

"At the time I joined KAUST, it had the best state-of-the-art facilities and [it] still does," he noted. "Also, the availability of funding to tackle interesting research problems was an important factor in my decision to [come to] KAUST. Over the years, this risk-taking attitude has helped me to distinguish myself from other successful researchers in my field."

"I was also attracted to KAUST because of its research center/degree program structure that allows for easy interactions between researchers that work in completely different fields," Bagci continued. "Ten years ago, when I first started at KAUST, it was the only institution that offered such a collaborative research environment.

"Ten years later, it may not be the only one, but it is one of the few global institutions that has provided a very fruitful interdisciplinary collaborative base for its researchers."

Advancing the growing discipline of computational electromagnetics

Bagci and his group at KAUST research the interdisciplinary field of computational electromagnetics (CEM). CEM fuses elements of electrical engineering, physics and applied mathematics and computational sciences to enable accurate, efficient and robust numerical modeling and characterization of real-life electromagnetic, optical and photonic systems.

"From this perspective, CEM complements—and more importantly enables—other branches of electromagnetics, optics and photonics that often rely on experimentation," he explained.

"Numerical design and characterization of real-life electromagnetic, optical and photonic systems [are]...challenging task[s] because computation domains are huge; the frequency of operation has a large dynamic range; devices are geometrically intricate with dimensions varying by orders of magnitude; and design frameworks require many repetitions of simulations with different parameters," Bagci noted.

Associate Professor Hakan Bagci was motivated to join the University because of its state-of-the-art facilities in the Supercomputing Core Lab. File photo.

 ​

"These requirements translate into large matrix systems with millions (if not billions) of unknowns to be solved for multi-scale discretization and ill-conditioned equations followed by prohibitively long execution times," he continued. "My research group at KAUST is trying to address these challenges by developing novel efficient and accurate algorithms and numerical schemes for solving integral/differential forms of Maxwell equations."

Bagci is excited to explore two main future research directions—the use of numerical modeling tools in data science and the development of multiphysics simulation techniques.

"Machine learning techniques often rely on the generation of big data sets. One way of doing this is through taking real measurements," he said. "However, for many applications, this might be costly or even impossible due to the long duration of time the measurements require. In such cases, simulation tools can be used to replace experiments. We are looking into ways of doing this in the field of electromagnetics."

"Engineering problems often involve more than one simultaneously happening physical phenomenon," Bagci continued. "For example, an electrical engineer who wants to design a chip has to know about electromagnetics, semiconductor physics, optoelectronics and even thermal analysis. Formulating and implementing efficient and accurate numerical schemes capable of modeling/simulating strongly coupled multiphysics phenomena is not trivial. My group at KAUST is looking into developing such techniques."

Related stories:

• Perfect absorption graphene style
• Bringing signals into phase
• A broader bandwidth for electrical devices