Stories – Technologies

Technologies – Expert view

Innovation of tomorrow: faster, more energy-efficient, networked.

Watch the video 
About Luc Van den hove

Luc Van den hove is President and CEO of imec, a leading research hub for the semiconductor industry. He joined imec in 1984, starting his research career in the field of silicide and interconnect technologies and went on to become COO and EVP of the silicon process and device technology division before becoming CEO in 2009. He is a professor of electrical engineering at KU Leuven, Belgium, where he also received his PhD in electrical engineering, and has authored or co-authored more than 200 publications and conference contributions. He is a passionate speaker on technology trends and applications for nanoelectronics at major conferences.

Imec’s Luc Van den hove has spent his entire career working to accelerate collaboration and innovation in the semiconductor space. “When I entered this field,” he says, “the industry was just transitioning from microelectronics to nanoelectronics. The first desktop PCs were going on desks.” Van den hove knew it was the beginning of a revolution, and he wanted to be on the front lines.

Thirty-seven years later, as the industry once again comes up against the limits of Moore’s Law, Van den hove is as enthusiastic as he was on his first day at imec.

The industry is seeing unprecedented demand for semiconductors due to a confluence of factors. One is the sheer amount of data being created. Another is the growth of new and expanded applications such as in cars, which now have hundreds of sensors each, and new healthcare innovations such as precision medicine.

Other areas driving semiconductor demand are the Internet of Things (IoT), artificial intelligence (AI), smart energy and smart cities. Together, these trends will lead demand for chips to double within the next ten years, Van den hove predicts, as tech innovation permeates virtually every sector of the world economy.

Van den hove foresees a future where AI calculations move away from the cloud. Transferring the data not only consumes too much energy, but also too much time. “We’re going to see a lot more of what we like to call distributed AI, where we bring intelligence much closer to the sensors." says Van den hove.

Perhaps the best example to imagine is autonomous driving. You can’t have latency when that car needs to stop suddenly. This means more ICs in cars which, Van den hove says, “are becoming computers on wheels.”

“We need technologies that consume less power, and we need to develop those technologies in a more sustainable way.”

New technologies for less energy consumption

A key challenge for the industry is that the exponential increase in data that we all generate and consume every day also results in exponential growth in the amount of energy consumed, which at some point would outpace the amount of energy available on Earth. “We need technologies that consume less power, and we need to develop those technologies in a more sustainable way,” says Van den hove. “At imec, we’re doing a lot of work to reduce what’s needed to produce.”

There are two trends Van den hove is watching on the manufacturing side: scaling and 3D heterogeneous integration.

Scaling is about reducing feature sizes. However, Van den hove says, “Scaling alone will not provide the required improvement needed as we evolve towards ever more complex circuitry. And that’s why a lot of focus now goes into exploring the third dimension and stacking those chips on top of each other. This is what we call 3D heterogeneous integration.” Using the third dimension is one of the advancements that will allow the industry to keep Moore’s Law in force, says Van den hove.

New plasma processing technologies are playing a crucial role in this. “We not only need to finetune the plasma processes to do the patterning,” says Van den hove “but once we explore the third dimension more and more, the number of sequential processing steps increases, and this will also require a lot more plasma processing.”

“We are bringing together key players and moving from a sequential to a network innovation model. This accelerates innovation.”

Industry collaboration accelerates progress

Collaboration in the supply chain is crucial in working toward these goals and is a primary aim for imec. Collaboration is also important because the industry is hitting the wall of physical limits.

Another dimension of collaboration is making sure suppliers have five to ten years’ notice of what will be needed to produce ICs of the future – for instance, new materials such as gallium nitride and silicon carbide, and new plasma processing equipment.

“In the past, innovation was sequential, and that took time. We are bringing together key players and moving from a sequential to a network innovation model. This accelerates innovation,” says Van den hove.