Research Program - Overview
Already today, impairments of the underlying semiconductor technology (e.g., probabilistic device behavior) have to be compensated by robust system designs compromising the achievable performance increases. Thus, high impact breakthroughs that are based on new materials can only be achieved if complete system solutions are considered: from materials all the way to large-scale integrated processing systems. We therefore comprehensively address all three Abstraction Layers: Materials & Functions, Devices & Circuits, and Information Processing.
It is the unique approach of our Cluster of Excellence to drive forward several different technology candidates, inspired by new materials, to a state where information processing becomes possible and to prepare for their integration in heterogeneous large-scale processing systems. We refer to our research areas as Research Paths to highlight the exploratory dynamic character in search of breakthroughs. By implementing this more shots on goal approach, we want to maximize the chances for high impact technological breakthroughs and will greatly benefit from the various possibilities for cross-fertilization between the Paths.
The Silicon Nanowire Path (Silicon NW) follows the most “conventional” approach by virtue of using silicon. Nevertheless, the potential electronic properties of these nanowires go far beyond current silicon technologies. Besides very beneficial electronic properties, silicon nanowires can be configured to change transistors between p- and n-type dynamically. Potentially, complex functionality can thus be implemented with much smaller numbers of devices. In this context, the design of novel and fault tolerant computing algorithms that make use of the transistors’ multi-functionality is also targeted. Furthermore, silicon nanowires will be explored as a selective sensor platform for biomolecules, opening a completely new application domain.
Carbon is clearly an outstanding candidate for advancing electronics beyond today’s boundaries. Within the Carbon Path we will explore carbon nanotubes (CNTs) as the currently most advanced form of carbon for use in electronic systems, with wireless communication as the main application. CNTs will also be a vehicle to extend the understanding of other carbon structures such as graphene at a later stage.
Organic materials are already successfully used in displays based on organic light-emitting diodes. However, there are still many obstacles for using organic electronics for information processing. Despite its inferiority to CMOS in terms of, e.g., speed, organic electronics systems could enable completely new applications in e.g., wearable electronic systems at very low cost. The goal of the Organic and Polymer Path is to overcome some of today’s major limitations of organic electronics and thereby enable new organic information processing systems.
A more unconventional approach is followed by the Biomolecular-Assembled Circuits Path (BAC). Here, complex biomolecular structures will serve as templates for metallization and functionalization. By using e.g., DNA origami, biomolecular nano-structures have the potential to build high quality inductors and other 3D circuit elements,which are difficult to construct with today’s planar silicon technologies. Through self-assembly, these structures can potentially be synthesized in a cost effective and highly parallel manner. By complementing silicon-based electronics, systems far beyond current boundaries could be enabled.
The Chemical Information Processing Path(CIP) uses chemical properties of substances (composition, physical state, concentration, etc.) as carriers of information. The unconventional approach is to process this chemical information in a controllable, integrated, massively parallel, microfluidic environment with various functional composition elements. Implemented with CMOS, this technology has the potential to revolutionize many processes involving feedback loops of chemical information (e.g., the synthesis of matter with designed properties) that are nowadays slow and manual and in doing so, open up completely new application domains.
The goal of the three System-Oriented Paths is to pave the way towards highly-efficient information processing on future augmented CMOS systems integrating heterogeneous solutions created by all Paths above.
The future technologies from the Material-Inspired Paths will lead to augmented CMOS systems with wildly varying properties and the potential to revolutionize the electronic systems landscape. The role of the Orchestration Path is to prepare the rapid and efficient implementation of these heterogeneous systems by addressing adaptation inflexibilities of current hard-and software designs. The goal of this Path is an automatic adaption of applications and the underlying systems software to new heterogeneous CMOS and augmented CMOS systems with minimal, ideally no, manual changes and, in particular, without losing the benefits of the underlying technologies due to additional complexities of traditional approaches.
Today, reliability issues already lead to diminishing performance returns when transitioning to smaller CMOS gate lengths. Soon the costs of traditional resilience mechanisms will cancel most of the benefits gained from transitioning to a new technology. The goal of the Resilience Path is to keep the costs of resilience as low as possible by focusing on flexible, application-specific, adaptive resiliency mechanisms. Reliable information processing with unreliable and adjustable components will be researched, taking into account the projected heterogeneity of future systems and the fault characteristics of new materials-inspired technologies.
The DFG Collaborative Research Center (CRC/SFB 912) "Highly Adaptive Energy Efficient Computing(HAEC)" complements this cluster as Path H. HAEC focuses on large-scale, multi-chip computing platforms with disruptive wireless and optical inter-chip interconnects and on hardware/software adaptation methods for a new quality of energy-efficient computing. The understanding generated here can have major impacts on the other System-Oriented Paths and will be challenged in a much wider context. E.g., new devices from the Materials-Inspired Paths could be integrated into HAEC’s computing platform. Therefore, the CRC is scientifically linked to cfAED, even though it is organizationally independent (see CRC 912).
Finally, it is the goal of the Biological Systems Path to apply new theoretical approaches to better understand the often surprising behavior of biological systems. The motivation for gaining new insight into information processing of biological systems is to discover whether biology offers interesting solutions for technical problems. This will be achieved by close cooperation between engineers and biologists in analyzing selected parts of biological systems. As such, this Path has the potential to inspire new solutions in all other Paths, especially at the circuits and information processing levels.