Issue link: https://iconnect007.uberflip.com/i/1299286
14 PCB007 MAGAZINE I OCTOBER 2020 there are many levels of IoT out there. It's go- ing to be everywhere. High-performance OEMs are going to be the first adopters. Autonomous vehicles are driving high levels of integration. When you cannot get the ca- pability in one device, you have to integrate more. Cellphones are becoming more powerful than our desktop or laptop in a smaller device. Miniaturization is happening. If you have an iPhone, there's a 3D IC in it. If you have any of the later ones, they're already stacked devices. We are already using heterogeneous integra- tion in even consumer applications. People are putting antennas on the same devices for ma- ny of these applications. Andy Shaughnessy: What challenges do you think a CAD manager or PCB designer should know? Horner: They need to maybe open their vo- cabulary horizon. Thermal is becoming criti- cal. We used to put heat sinks on a package or have all this heating and cooling and the fans in a system to cool the different parts. Warp- age is not limited to the package. The boards are warping because of the heat; they're de- laminating devices off of the board. The HIR provides a map of all the potential issues that will surface. Maybe it's focusing more on the package level right now, but it's affecting the board level, too. Sooner or later, the package will become what the old PCB was. Again, there are only so many of these 100- by 100-millimeter package parts these boards can handle in terms of the mechanical stress or the thermal impact. There may be multiples of these 100- by 100-millimeter square devices connected on one board. Is back drilling go- ing to do it? No. Maybe blind vias are better. Those are some of the things that I'm thinking about on a high-speed side that are already be- ing raised by a lot of high-end standards. Holden: Who's going to create some of the tools that collaborate and lead that connection be- tween these isolated silos? Is it going to be the OEMs because they want the final product, university students working on their Ph.D.s, or top EDA companies, which will begin as start- ups and then be bought by large EDA compa- nies and incorporated into their tools? Figure 2: Five levels of autonomous driving and one example of a roadmap to 2025. (Source: HIR 2019)