Abstract:

The next decade will likely see the arrival of many autonomously moving devices. From humanoid-like robots that walk and talk, to self-driving cars, and to delivering drones.

All enabled by sensors (position, movement, light, chemical),  actuators (motors, levers, displays), and local as well as cloud-based intelligence. Most of these based on semi-conductor devices and technology.

Movement requires force, work and power, and the semiconductor devices that can deliver and transform sufficient energy are built using very different dimensions and  process integration than the digital electronics used for computing and control. Bringing these two worlds together in small form factors requires heterogeneous integration or packaging, called modules. Recent progress in the power / volume, efficiency, and intelligence of modules has accelerated through much progress in the active devices, but also from substrate and package capabilities. Scaling power devices for lower cost and higher efficiency has been accomplished with trench based architectures, super-junction doping profiles, and ultra-thin wafer processes. Recent generations of IGBTs and power MOSFETs achieve on resistances and operating frequencies at tighter densities than ever, leading to much discussion about their ultimate limits, not unlike the prognostications in advanced logic technology. And if Silicon cannot be extended further, there are the wide bandgap materials of GaN and SiC to extend voltage, speed, or temperature.  Especially promising is the system level benefit of higher frequency operation. With smaller coils and capacitors, much smaller form factors can be achieved;  all eseentisl factors for autonomous mobility.

Co-packing all these unique components is enabled by progress in substrates with thinner, finer interconnect, yet with adequate heat dissipation from better materials choices. While the IQ of intelligent power modules has increased cost / performance in many fields, nowhere has the progress of 3D integration been more impressive than in image sensors. Here, use of wafer level stacking allows better optimization of the sensor layers, the analog transfer functions, and the digital image processing. Add to this the wafer level processes for lenses and focus techniques, and miniature imaging devices are a common reality, dealing now with extreme dynamic range and robustness requirements from the automotive world. The required intelligence for decision making at highway speeds is core to realizing the autonomous driving experience.

In this talk, we will cover the range of technical trade-offs encountered in scaling power devices, realizing the potential of wide bandgap materials through adjusted circuit topologies, and the challenges of heterogeneous packaging and integration. At last, after many decades of exponential progress in logic and memory technologies, the “real world” devices of power handling and sensor functions are jointly enabling another wave of electronics progress, namely all things autonomously moving.

Dr. Hans Stork joined ON Semiconductor in 2011 as Senior Vice President and Chief Technology Officer (CTO) responsible for the company’s product research and development.

With nearly three decades of experience in technology and product development, Dr. Stork has held various leadership positions including Group Vice President and CTO of the Si Systems Group at Applied Materials, and Senior Vice President and CTO of Texas Instruments and director of Si Technology Development at TI. During his career, Dr. Stork has worked with some of the semiconductor industry’s leading innovators including Hewlett Packard and IBM Research.

Dr. Stork holds a PhD in Electrical Engineering from Stanford University, and an Electrical Engineering Ingenieur (Cum Laude) from Delft University in the Netherlands. He has authored numerous scientific articles and papers, and served as a board director for Sematech and the Semiconductor Research Corporation (SRC) and is a member of the Scientific Advisory Board for IMEC.