ROBUST Designing Robust Nanoelectric Systems

Motivation

Electronic systems enable many German key industries to be successful – whether this be the automobile industry, medical technology, mechanical engineering or, of course, the whole electro and information technology industry. Furthermore, electronic systems are contributing to an ever-increasing extent, whether it be with infrastructure or with consumer products, not only to making our modern society more efficient but also to safeguarding people’s health. This is very evident in medical electronics.
With new electronic systems in the automobile industry, which have been introduced recently or are currently being explored in research and development, (adaptive cruise control; lane recognition and lane keeping support; fatigue warning system; and ultimately: autonomous vehicles) the vision “of zero road deaths and serious injuries” (“Vision Zero“) begins to move within tangible distance. Today’s world is unimaginable without products based on nanoelectronic system components. In future, the success of the national economies will depend more and more on mastering the continually increasing use of electronics ahead of the competition. Examples are given for the automobile industry in.
Electronic systems to process and control information with vital importance cannot just be developed like consumer products. Today integrated circuits already contain up to a billion transistors in which each transistor may fail. This complexity will continue to increase exponentially, thus allowing attractive new applications. One of the challenges facing us is to consider during the design phase to prepare robust operation modes . It should be measurable and therefore comparable. Not only has progress in processing resulted in ever smaller structure dimensions but also the resulting systems are immensely complex.

With the BMBF project ROBUST (grant no. 01 M 3087) new innovative design techniques are being explored for German industry in order to be able to produce electronic systems for robust operation cost-effectively in 5-10 years’ time. In order to achieve this the project is researching robustness data on the one hand to be able to assess robustness and on the other hand develop design methodology to specifically improve this robustness. ROBUST can therefore, on the basis of robustness data, be used to quantify and specifically improve robustness in operation throughout the whole lifetime in electronic systems. Meanwhile ROBUST provides German industry with new design methods so that industry can adapt the new methods in its designing processes to industrial requirements. To do this ROBUST encourages dialogue with industry in order to support it in increasing the robustness of electronic systems.

Technologies

With ROBUST the robustness of an SoC is already considered during the design process since this is the time when it can most effectively be increased. Architectural measures are already taken on the system level for reducing stress, for stress tolerance and increasing reliability as measures applied to deeper abstraction levels only have a local effect and compensating time controlled variabilities or faulty behavior is difficult to put into practice. Furthermore the aim is to bring about measures which increase robustness on the system level through the close relationship to applications. The reason for this is that intelligent compensation measures, unlike simple structural redundancy processes, require less area and are therefore less expensive to achieve. Conversely, meaningful system considerations can only be made if technology-specific influence factors on the robustness of a system can be described adequately by abstract component models which are applicable on the system level. Additionally, various types of platform components are included in the robustness considerations. In this way the robustness of digital modules will be improved in the same way as with analog and mixed-signal modules.

Persons
Publications
Robustness evaluation of embedded software systems

Lu, Weiyun and Radetzki, Martin and Metzdorf, Malte and Helms, Domenik and Nebel, Wolfgang; 005 / 2011

MoRV: Modelling Reliability under Variability

Malte Metzdorf, Reef Eilers, Domenik Helms, Wolfgang Nebel, Kay-Uwe Giering, Roland Jancke, Gerhard Rzepa, Tibor Grasser, Markus Karner, Praveen Raghavan, Ben Kaczer, Dan Alexandrescu, Adrian Evans, Gunnar Rott, Peter Rotter, Hans Reisinger, Wolfgang Gustin; SELSE 12 proceedings; 03 / 2016

Behavioral-Level Thermal- and Aging-Estimation Flow

Rosinger, Sven and Metzdorf, Malte and Helms, Domenik and Nebel, Wolfgang; Test Workshop (LATW), 2011 12th Latin American; 03 / 2011

Design methodology for a self-healing signal processing unit on the example of an FIR filter

[Metzdorf, Malte and Eilers, Reef and Helms, Domenik and Nebel, Wolfgang]; 005 / 2012

Robustness evaluation of embedded software systems

[Lu, Weiyun and Metzdorf, Malte and Helms, Domenik and Radetzki, Martin and Nebel, Wolfgang]; 005 / 2011

Bewertung und Optimierung der Robustheit mikroelektronischer Systeme

[Metzdorf, Malte and Helms, Domenik]; 009 / 2012

Closing the gap between technology and application needs

Nebel, Wolfgang and Helms, Domenik and Grüttner, Kim and Oppenheimer, Frank; 05 / 2013

Abstracting TCAD aging models above the circuit level

Malte Metzdorf, Domenik Helms, Reef Eilers, Wolfgang Nebel; DATE - Design, Automation, and Test in Europe; 03 / 2015

RT Level Timing Model for Aging Prediction

Nils Koppaetzky, Malte Metzdorf, Reef Eilers, Domenik Helms, Wolfgang Nebel; DATE; 03 / 2016

Analysis of NBTI Effects on High Frequency Digital Circuits

Ahmet Unutulmaz, Domenik Helms, Reef Eilers, Malte Metzdorf, Ben Kaczer, Wolfgang Nebel; DATE; 03 / 2016

Leakage models for high level power estimation

Domenik Helms ; Reef Eilers ; Malte Metzdorf ; Wolfgang Nebel; IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems; 11 / 2017

Partners
Edacentrum GmbH
www.edacentrum.com
TU München, Fakultät für Informatik, Lehrstuhl IV: Software & Systems Engineering
www4.informatik.tu-muenchen.de
Universität Stuttgart
FZI Forschungszentrum Informatik
www.fzi.de
Goethe Universität Frankfurt
www.uni-frankfurt.de

Duration

Start: 31.03.2009
End: 30.03.2012