Semiconductor IP News and Trends Blog

Imec Talks Boundaries, Innovation, and IP at Semicon West

Crossing interdisciplinary boundaries, 3D stacked die collaboration, open innovation and IP challenges were the topics of an Imec interview at Semicon West.

How can the success of the semiconductor business model and technology be translated into other markets? What is a proven way to encourage multi-disciplinary innovation and IP? To answer these questions and others, I spoke with Luc Van den hove, the President and CEO of Imec – a major nanotechnology R&D leader located in Belgium. What follows is an edited version of our conversation.  – JB

Blyler: At the Imec Technology Forum (ITF) keynote during Semicon West, you said that innovation occurs at the boundary of disciplines and noted Imec’s position at that interdisciplinary crossing in the semiconductor world. The challenge with interdisciplinary projects is getting everyone coordinated and working as a team. This is typically the role of the systems engineer or integrator. How does Imec ensure that multi-disciplinary teams work together?

Van den hove: We know the revolution that semiconductor technology has realized in the ICT (Information and Communications Technology) world. My point was that to address new markets you need to bring together disciplines from completely different fields like the electronics discipline with growing opportunities in the biomedical field. So we have to bring together people with biotech backgrounds, medical doctors and engineers, all of which speak a completely different language. That is why we have set up common labs where these people can work together in one environment. We believe this approach is the most effective way to have those different communities and disciplines work together.

Blyler: One of the ongoing challenges in the semiconductor world is optimizing the collaboration between chip designers (EDA-IP), equipment manufacturers, foundries and more recently the packaging vendors. This collaboration will become critical with the move to three-dimensional (3D) stack dies. What roll does Imec play in this collaboration?

ETNA chip, a 3D chip integrating a commercial DRAM chip on top of a 25µm-thick logic IC using through-silicon-via (TSV) technology developed at Imec.

Van den hove: We are trying to set up a 3D ecosystem where all the key players can come together. Six to seven years ago when this topic first gained attention, it was quite difficult to know which direction 3D integration would take. Plus you have involvement from a lot of (different) players. For example, the equipment community was used to working in the packaging domain but with the move to 3D will have to interface with the foundries. Further, the EDA players that have to develop specific software to design architectures in 3D. And the foundries and the fabless companies will need to understand the effect 3D integration will have on their manufacturing and design processes (respectively).

This is why we brought together all the key players in the 3D domain. At Imec, we merged our packaging with our interconnect team to help narrow down the integration routes that are of interest to reach a common platform, such as through silicon technology which has become a mainstream integration routes for 3D. We could only achieve this by bringing together all the key players: foundries; OSATs; equipment suppliers; material companies and even the end users like the fabless companies.

Blyler: The goal is not to create standards but promote early engagement on these problems. Is that correct?

Van den hove: Indeed. Each of the two major worlds – packaging and the chip-foundry players – have different ways of doing R&D and of performing integrations. So we must bring together these two worlds to find out the best integration schemes that will really address the key challenges on an architecture level.

Blyler: Let’s switch focus just a bit. Intellectual property (IP) trends such as patent trolling and the like seem to have a negative affect on innovation. Imec promotes an open innovation approach that is different from consortium arrangements and may circumvent IP patent battles. What’s new on the IP front at Imec?

Van den hove: The business model at Imec is a very flexible one. It is indeed open innovation in the sense that we bring together a lot of the key players of the ecosystem to work on common challenges. But that doesn’t mean that all the IP and the results are shared among all the players. The pre-competitive IP is shared among the key contributors. But what is pre-competitive to one part of the value chain may be competitive for another part. Our model is optimized to the various segments of the value chain. For example, when we talk to material or equipment suppliers, some of those things may be competitive whereas for the users of the equipment it may be pre-competitive since it is for the next (process) note. So we have adapted our model to be something that is much more sophisticated than a consortium model in which everything is shared amongst all the members.

We also have the possibility of developing proprietary IP on certain parts of the program – – in addition to the common platform that we set up and share among many partners. Any proprietary IP is something that we agree to up-front with the partners so it doesn’t interfere with the rest of the program. The basic message is that our model is tuned toward creating a win-win relationship with each of our partners. It is not necessarily a model to share everything with everybody. That wouldn’t work.

Related Articles:

• Avoid IP Patent Battles with Open Innovation?