Différences entre les versions de « Technologies pour la CAO »
| Ligne 67 : | Ligne 67 : | ||
| − | [[En: | + | [[En:Main_Page]] |
Version du 13 septembre 2016 à 16:27
L'intégration de systèmes toujours plus complexes requière des méthodologies et outils de conception adaptés. En effet, les contraintes fonctionnelles et technologiques sont en constante augmentation alors que les exigences ainsi que le temps de mise sur le marché sont toujours plus serrées. Dans le but de répondre à ces contraintes les approches classiques doivent revues et adaptées. Elles doivent tenir compte de concepts complémentaires tels que le formalisme augmenté, la "réutilisabilité" et la capitalisation des savoirs.
L'équipe SMH développe des modèles compacts et des outils de design multi-physiques dédiés aux systèmes instrumentaux à architectures analogiques et mixtes. Nous développons de nouvelles approches de conception pour les systèmes multidisciplinaires en nous basant sur la spécification formelle et les langages de description matériel. Ceci est réalisé en étendant le champ d'application de techniques de mathématiques, développées dans d'autre buts tel que l'arithmétique des intervalles. L'objectif est d'améliorer les vitesses de calcul des simulateurs tout en restant au plus proche du comportement physique des dispositifs avancés.
Modélisation compacte
This research activity mainly focuses on the compact modeling of new fundamental devices such as ultimate transistors and also micro or nanoscale sensors developed within the team.
A large part of this activity has been dedicated to ultimate FET transistors modeling. For over 20 years now, our team has been closely working on that topic with teams from EPFL (Lausanne, Switzerland) and also URV (Tarragone, Spain). We carried out compact models for FinFET and multi-gate (featuring two, three, four gates and GAA) MOS devices. We have also been working on the modeling of junction-free nanowire-based FETs, which display promising performance in regard of future high-end IC technologies.
Sensors modeling is another part of our activity and represents a necessary step for optimizing integrated instrumentation systems. Within that framework we developed models for magnetic sensors based on the Hall effect (HHD and VHD) and charge deflection (CHOPFET).
More recently we started new research projects on the modeling of chemical sensors. We have been specifically studying two kinds of sensors, carbon nanotube FET (CNTFET) on one hand and ion sensitive FET (ISFET) on the other. Both devices can be potentially functionalized in order to be used as biosensors.
Related projects
The principal projects related to the compact modeling activities are:
- FP7 COMON, ended 2012, "Modeling of multi-gate SOI transistors and of junction-free nanowire-based transistors"
- XYZ-IRM project, "Development of a dedicated system for active minimally-invasive surgery tools tracking in IRM environment"
- ANR CAPTEX, 2010-2013, "Modeling of CNTFET-based explosive gas sensors and development of their dedicated fully integrated conditioning electronics"
- industrial project in collaboration with BURKERT company, "Virtual prototyping of Lab-on-Chip (LoC) for pollutant detection"
High-level multiphysics modeling and simulation tools
In order to be able to meet the ever growing requirements in IT and mobile devices, the IC manufacturers gradually turn to 3D integration, which consists in stacking chips thus forming complex single-package systems. Such an approach allows to look into the opportunity of new solutions to stick to Moore's law or even achieve "more than Moore" performance, but it also come with several shortcomings. In particular, evacuating the heat produced within an integrated chip is particularly difficult and induces an increase of the average operating temperature and also hotter spots. Besides being responsible for higher power consumption, these thermal issues also induce additional mechanical constraints due to material expansion and have overall negative impact on chips' reliability.
As it is the aging mechanisms that occur in integrated circuits (electromigration NBTI,...) are all the more important with high temperature. In order to minimize these harmful effects, it has now become essential to take thermal and mechanical characteristics at the early stage of IC development.
In that context, since 2009 we have been working on the development of a multiphysics (i.e. electrothermal and mechanical) simulator, embedded into one of industry's standard CAD tool, namely CADENCE®. Now, our team is in possession of an experimentally validated operational tool dedicated to simulating electrothermal and mechanical behavior in the IC engineering flow. This enables high accuracy circuit behavior simulation and should ultimately allow accurate forecasting of an integrated system's lifetime, i.e. prior to failure.
Related projects
The principal projects related to the compact modeling activities are:
- ANR 3D-IDEAS, ended 2012, "Integration and 3D conception technology for imaging systems and applications"
