• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 28
  • 6
  • 5
  • 3
  • 1
  • Tagged with
  • 52
  • 52
  • 12
  • 10
  • 10
  • 8
  • 7
  • 7
  • 6
  • 6
  • 6
  • 6
  • 6
  • 5
  • 5
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
51

Modeling and Analysis of Large-Scale On-Chip Interconnects

Feng, Zhuo 2009 December 1900 (has links)
As IC technologies scale to the nanometer regime, efficient and accurate modeling and analysis of VLSI systems with billions of transistors and interconnects becomes increasingly critical and difficult. VLSI systems impacted by the increasingly high dimensional process-voltage-temperature (PVT) variations demand much more modeling and analysis efforts than ever before, while the analysis of large scale on-chip interconnects that requires solving tens of millions of unknowns imposes great challenges in computer aided design areas. This dissertation presents new methodologies for addressing the above two important challenging issues for large scale on-chip interconnect modeling and analysis: In the past, the standard statistical circuit modeling techniques usually employ principal component analysis (PCA) and its variants to reduce the parameter dimensionality. Although widely adopted, these techniques can be very limited since parameter dimension reduction is achieved by merely considering the statistical distributions of the controlling parameters but neglecting the important correspondence between these parameters and the circuit performances (responses) under modeling. This dissertation presents a variety of performance-oriented parameter dimension reduction methods that can lead to more than one order of magnitude parameter reduction for a variety of VLSI circuit modeling and analysis problems. The sheer size of present day power/ground distribution networks makes their analysis and verification tasks extremely runtime and memory inefficient, and at the same time, limits the extent to which these networks can be optimized. Given today?s commodity graphics processing units (GPUs) that can deliver more than 500 GFlops (Flops: floating point operations per second). computing power and 100GB/s memory bandwidth, which are more than 10X greater than offered by modern day general-purpose quad-core microprocessors, it is very desirable to convert the impressive GPU computing power to usable design automation tools for VLSI verification. In this dissertation, for the first time, we show how to exploit recent massively parallel single-instruction multiple-thread (SIMT) based graphics processing unit (GPU) platforms to tackle power grid analysis with very promising performance. Our GPU based network analyzer is capable of solving tens of millions of power grid nodes in just a few seconds. Additionally, with the above GPU based simulation framework, more challenging three-dimensional full-chip thermal analysis can be solved in a much more efficient way than ever before.
52

SPICE Modeling of TeraHertz Heterojunction bipolar transistors / Modélisation compacte des transistors bipolaires fonctionnant dans la gamme TeraHertz

Stein, Félix 16 December 2014 (has links)
Les études qui seront présentées dans le cadre de cette thèse portent sur le développement et l’optimisation des techniques pour la modélisation compacte des transistors bipolaires à hétérojonction (TBH). Ce type de modélisation est à la base du développement des bibliothèques de composants qu’utilisent les concepteurs lors de la phase de simulation des circuits intégrés. Le but d’une technologie BiCMOS est de pouvoir combiner deux procédés technologiques différents sur une seule et même puce. En plus de limiter le nombre de composants externes, cela permet également une meilleure gestion de la consommation dans les différents blocs digitaux, analogiques et RF. Les applications dites rapides peuvent ainsi profiter du meilleur des composants bipolaires et des transistors CMOS. Le défi est d’autant plus critique dans le cas des applications analogiques/RF puisqu’il est nécessaire de diminuer la puissance consommée tout en maintenant des fréquences de fonctionnement des transistors très élevées. Disposer de modèles compacts précis des transistors utilisés est donc primordial lors de la conception des circuits utilisés pour les applications analogiques et mixtes. Cette précision implique une étude sur un large domaine de tensions d’utilisation et de températures de fonctionnement. De plus, en allant vers des nœuds technologiques de plus en plus avancés, des nouveaux effets physiques se manifestent et doivent être pris en compte dans les équations du modèle. Les règles d’échelle des technologies plus matures doivent ainsi être réexaminées en se basant sur la physique du dispositif. Cette thèse a pour but d’évaluer la faisabilité d’une offre de modèle compact dédiée à la technologie avancée SiGe TBH de chez ST Microelectronics. Le modèle du transistor bipolaire SiGe TBH est présenté en se basant sur le modèle compact récent HICUMversion L2.3x. Grâce aux lois d’échelle introduites et basées sur le dessin même des dimensions du transistor, une simulation précise du comportement électrique et thermique a pu être démontrée.Ceci a été rendu possible grâce à l’utilisation et à l’amélioration des routines et méthodes d’extraction des paramètres du modèle. C’est particulièrement le cas pour la détermination des éléments parasites extrinsèques (résistances et capacités) ainsi que celle du transistor intrinsèque. Finalement, les différentes étapes d’extraction et les méthodes sont présentées, et ont été vérifiées par l’extraction de bibliothèques SPICE sur le TBH NPN Haute-Vitesse de la technologie BiCMOS avancée du noeud 55nm, avec des fréquences de fonctionnement atteignant 320/370GHz de fT = fmax. / The aim of BiCMOS technology is to combine two different process technologies intoa single chip, reducing the number of external components and optimizing power consumptionfor RF, analog and digital parts in one single package. Given the respectivestrengths of HBT and CMOS devices, especially high speed applications benefit fromadvanced BiCMOS processes, that integrate two different technologies.For analog mixed-signal RF and microwave circuitry, the push towards lower powerand higher speed imposes requirements and presents challenges not faced by digitalcircuit designs. Accurate compact device models, predicting device behaviour undera variety of bias as well as ambient temperatures, are crucial for the development oflarge scale circuits and create advanced designs with first-pass success.As technology advances, these models have to cover an increasing number of physicaleffects and model equations have to be continuously re-evaluated and adapted. Likewiseprocess scaling has to be verified and reflected by scaling laws, which are closelyrelated to device physics.This thesis examines the suitability of the model formulation for applicability to production-ready SiGe HBT processes. A derivation of the most recent model formulationimplemented in HICUM version L2.3x, is followed by simulation studies, whichconfirm their agreement with electrical characteristics of high-speed devices. Thefundamental geometry scaling laws, as implemented in the custom-developed modellibrary, are described in detail with a strong link to the specific device architecture.In order to correctly determine the respective model parameters, newly developed andexisting extraction routines have been exercised with recent HBT technology generationsand benchmarked by means of numerical device simulation, where applicable.Especially the extraction of extrinsic elements such as series resistances and parasiticcapacitances were improved along with the substrate network.The extraction steps and methods required to obtain a fully scalable model library wereexercised and presented using measured data from a recent industry-leading 55nmSiGe BiCMOS process, reaching switching speeds in excess of 300GHz. Finally theextracted model card was verified for the respective technology.

Page generated in 0.1183 seconds