Multicore Scheduling of Real-Time Irregular Parallel Algorithms in Linux

Face à estagnação da tecnologia uniprocessador registada na passada década, aos principais fabricantes de microprocessadores encontraram na tecnologia multi-core a resposta `as crescentes necessidades de processamento do mercado. Durante anos, os desenvolvedores de software viram as suas aplicações acompanhar os ganhos de performance conferidos por cada nova geração de processadores sequenciais, mas `a medida que a capacidade de processamento escala em função do número de processadores, a computação sequencial tem de ser decomposta em várias partes concorrentes que possam executar em paralelo, para que possam utilizar as unidades de processamento adicionais e completar mais rapidamente. A programação paralela implica um paradigma completamente distinto da programação sequencial. Ao contrário dos computadores sequenciais tipificados no modelo de Von Neumann, a heterogeneidade de arquiteturas paralelas requer modelos de programação paralela que abstraiam os programadores dos detalhes da arquitectura e simplifiquem o desenvolvimento de aplicações concorrentes. Os modelos de programação paralela mais populares incitam os programadores a identificar instruções concorrentes na sua lógica de programação, e a especificá-las sob a forma de tarefas que possam ser atribuídas a processadores distintos para executarem em simultâneo. Estas tarefas são tipicamente lançadas durante a execução, e atribuídas aos processadores pelo motor de execução subjacente. Como os requisitos de processamento costumam ser variáveis, e não são conhecidos a priori, o mapeamento de tarefas para processadores tem de ser determinado dinamicamente, em resposta a alterações imprevisíveis dos requisitos de execução. `A medida que o volume da computação cresce, torna-se cada vez menos viável garantir as suas restrições temporais em plataformas uniprocessador. Enquanto os sistemas de tempo real se começam a adaptar ao paradigma de computação paralela, há uma crescente aposta em integrar execuções de tempo real com aplicações interativas no mesmo hardware, num mundo em que a tecnologia se torna cada vez mais pequena, leve, ubíqua, e portável. Esta integração requer soluções de escalonamento que simultaneamente garantam os requisitos temporais das tarefas de tempo real e mantenham um nível aceitável de QoS para as restantes execuções. Para tal, torna-se imperativo que as aplicações de tempo real paralelizem, de forma a minimizar os seus tempos de resposta e maximizar a utilização dos recursos de processamento. Isto introduz uma nova dimensão ao problema do escalonamento, que tem de responder de forma correcta a novos requisitos de execução imprevisíveis e rapidamente conjeturar o mapeamento de tarefas que melhor beneficie os critérios de performance do sistema. A técnica de escalonamento baseado em servidores permite reservar uma fração da capacidade de processamento para a execução de tarefas de tempo real, e assegurar que os efeitos de latência na sua execução não afectam as reservas estipuladas para outras execuções. No caso de tarefas escalonadas pelo tempo de execução máximo, ou tarefas com tempos de execução variáveis, torna-se provável que a largura de banda estipulada não seja consumida por completo. Para melhorar a utilização do sistema, os algoritmos de partilha de largura de banda (capacity-sharing) doam a capacidade não utilizada para a execução de outras tarefas, mantendo as garantias de isolamento entre servidores. Com eficiência comprovada em termos de espaço, tempo, e comunicação, o mecanismo de work-stealing tem vindo a ganhar popularidade como metodologia para o escalonamento de tarefas com paralelismo dinâmico e irregular. O algoritmo p-CSWS combina escalonamento baseado em servidores com capacity-sharing e work-stealing para cobrir as necessidades de escalonamento dos sistemas abertos de tempo real. Enquanto o escalonamento em servidores permite partilhar os recursos de processamento sem interferências a nível dos atrasos, uma nova política de work-stealing que opera sobre o mecanismo de capacity-sharing aplica uma exploração de paralelismo que melhora os tempos de resposta das aplicações e melhora a utilização do sistema. Esta tese propõe uma implementação do algoritmo p-CSWS para o Linux. Em concordância com a estrutura modular do escalonador do Linux, ´e definida uma nova classe de escalonamento que visa avaliar a aplicabilidade da heurística p-CSWS em circunstâncias reais. Ultrapassados os obstáculos intrínsecos `a programação da kernel do Linux, os extensos testes experimentais provam que o p-CSWS ´e mais do que um conceito teórico atrativo, e que a exploração heurística de paralelismo proposta pelo algoritmo beneficia os tempos de resposta das aplicações de tempo real, bem como a performance e eficiência da plataforma multiprocessador.

Towards cryogenic liquid –vapor energy storage units for space applications

Dissertation to obtain the Doctoral degree in Physics Engineering

A system’s approach to cache hierarchy-aware decomposition of data-parallel computations

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Fishing the key biological components of the bacterium Geobacter metallireducens for optimal biotechnological applications

Os resultados apresentados no capítulo 2 foram incluídos no artigo Dantas JM, Campelo LM, Duke NEC, Salgueiro CA, Pokkuluri PR (2015) "The structure of PccH from Geobacter sulfurreducens – a novel low reduction potential monoheme cytochrome essential for accepting electrons from an electrode", FEBS Journal, 282, 2215-2231.

Switching mode power amplifier for bluetooth applications

Modern fully integrated transceivers architectures, require circuits with low area, low cost, low power, and high efficiency. A key block in modern transceivers is the power amplifier, which is deeply studied in this thesis. First, we study the implementation of a classical Class-A amplifier, describing the basic operation of an RF power amplifier, and analysing the influence of the real models of the reactive components in its operation. Secondly, the Class-E amplifier is deeply studied. The different types of implementations are reviewed and theoretical equations are derived and compared with simulations. There were selected four modes of operation for the Class-E amplifier, in order to perform the implementation of the output stage, and the subsequent comparison of results. This led to the selection of the mode with the best trade-off between efficiency and harmonics distortion, lower power consumption and higher output power. The optimal choice was a parallel circuit containing an inductor with a finite value. To complete the implementation of the PA in switching mode, a driver was implemented. The final block (output stage together with the driver) got 20 % total efficiency (PAE) transmitting 8 dBm output power to a 50 W load with a total harmonic distortion (THD) of 3 % and a total consumption of 28 mW. All implementations are designed using standard 130 nm CMOS technology. The operating frequency is 2.4 GHz and it was considered an 1.2 V DC power supply. The proposed circuit is intended to be used in a Bluetooth transmitter, however, it has a wider range of applications.

An evolutionary approach to the use of Petri net based models: from parallel controllers to HW/SW co-design

"A workshop within the 19th International Conference on Applications and Theory of Petri Nets - ICATPN’1998"

Parallel Architecture Prototype for 60 GHz High Data Rate Wireless Single Carrier Receiver

Nowadays a huge attention of the academia and research teams is attracted to the potential of the usage of the 60 GHz frequency band in the wireless communications. The use of the 60GHz frequency band offers great possibilities for wide variety of applications that are yet to be implemented. These applications also imply huge implementation challenges. Such example is building a high data rate transceiver which at the same time would have very low power consumption. In this paper we present a prototype of Single Carrier -SC transceiver system, illustrating a brief overview of the baseband design, emphasizing the most important decisions that need to be done. A brief overview of the possible approaches when implementing the equalizer, as the most complex module in the SC transceiver, is also presented. The main focus of this paper is to suggest a parallel architecture for the receiver in a Single Carrier communication system. This would provide higher data rates that the communication system canachieve, for a price of higher power consumption. The suggested architecture of such receiver is illustrated in this paper,giving the results of its implementation in comparison with its corresponding serial implementation.

Exploring Cross-layer power management for PGAS applications on the SCC platform

Technological limitations and power constraints are resulting in high-performance parallel computing architectures that are based on large numbers of high-core-count processors. Commercially available processors are now at 8 and 16 cores and experimental platforms, such as the many-core Intel Single-chip Cloud Computer (SCC) platform, provide much higher core counts. These trends are presenting new sets of challenges to HPC applications including programming complexity and the need for extreme energy efficiency.In this work, we first investigate the power behavior of scientific PGAS application kernels on the SCC platform, and explore opportunities and challenges for power management within the PGAS framework. Results obtained via empirical evaluation of Unified Parallel C (UPC) applications on the SCC platform under different constraints, show that, for specific operations, the potential for energy savings in PGAS is large; and power/performance trade-offs can be effectively managed using a cross-layerapproach. We investigate cross-layer power management using PGAS language extensions and runtime mechanisms that manipulate power/performance tradeoffs. Specifically, we present the design, implementation and evaluation of such a middleware for application-aware cross-layer power management of UPC applications on the SCC platform. Finally, based on our observations, we provide a set of recommendations and insights that can be used to support similar power management for PGAS applications on other many-core platforms.

Investigation of High Density Polyethylene Pipe for Highway Applications, HR-373, Phase 1, 1996

In the past, culvert pipes were made only of corrugated metal or reinforced concrete. In recent years, several manufacturers have made pipe of lightweight plastic - for example, high density polyethylene (HDPE) - which is considered to be viscoelastic in its structural behavior. It appears that there are several highway applications in which HDPE pipe would be an economically favorable alternative. However, the newness of plastic pipe requires the evaluation of its performance, integrity, and durability; A review of the Iowa Department of Transportation Standard Specifications for Highway and Bridge Construction reveals limited information on the use of plastic pipe for state projects. The objective of this study was to review and evaluate the use of HDPE pipe in roadway applications. Structural performance, soil-structure interaction, and the sensitivity of the pipe to installation was investigated. Comprehensive computerized literature searches were undertaken to define the state-of-the-art in the design and use of HDPE pipe in highway applications. A questionnaire was developed and sent to all Iowa county engineers to learn of their use of HDPE pipe. Responses indicated that the majority of county engineers were aware of the product but were not confident in its ability to perform as well as conventional materials. Counties currently using HDPE pipe in general only use it in driveway crossings. Originally, we intended to survey states as to their usage of HDPE pipe. However, a few weeks after initiation of the project, it was learned that the Tennessee DOT was in the process of making a similar survey of state DOT's. Results of the Tennessee survey of states have been obtained and included in this report. In an effort to develop more confidence in the pipe's performance parameters, this research included laboratory tests to determine the ring and flexural stiffness of HDPE pipe provided by various manufacturers. Parallel plate tests verified all specimens were in compliance with ASTM specifications. Flexural testing revealed that pipe profile had a significant effect on the longitudinal stiffness and that strength could not be accurately predicted on the basis of diameter alone. Realizing that the soil around a buried HDPE pipe contributes to the pipe stiffness, the research team completed a limited series of tests on buried 3 ft-diameter HDPE pipe. The tests simulated the effects of truck wheel loads above the pipe and were conducted with two feet of cover. These tests indicated that the type and quality of backfill significantly influences the performance of HDPE pipe. The tests revealed that the soil envelope does significantly affect the performance of HDPE pipe in situ, and after a certain point, no additional strength is realized by increasing the quality of the backfill.

Introducing massively parallel computers into operational weather forecasting

Numerical weather prediction and climate simulation have been among the computationally most demanding applications of high performance computing eversince they were started in the 1950's. Since the 1980's, the most powerful computers have featured an ever larger number of processors. By the early 2000's, this number is often several thousand. An operational weather model must use all these processors in a highly coordinated fashion. The critical resource in running such models is not computation, but the amount of necessary communication between the processors. The communication capacity of parallel computers often fallsfar short of their computational power. The articles in this thesis cover fourteen years of research into how to harness thousands of processors on a single weather forecast or climate simulation, so that the application can benefit as much as possible from the power of parallel high performance computers. The resultsattained in these articles have already been widely applied, so that currently most of the organizations that carry out global weather forecasting or climate simulation anywhere in the world use methods introduced in them. Some further studies extend parallelization opportunities into other parts of the weather forecasting environment, in particular to data assimilation of satellite observations.

FPGA-Based Control Design for Power Electronic Applications

Tehoelektoniikkalaitteella tarkoitetaan ohjaus- ja säätöjärjestelmää, jolla sähköä muokataan saatavilla olevasta muodosta haluttuun uuteen muotoon ja samalla hallitaan sähköisen tehon virtausta lähteestä käyttökohteeseen. Tämä siis eroaa signaalielektroniikasta, jossa sähköllä tyypillisesti siirretään tietoa hyödyntäen eri tiloja. Tehoelektroniikkalaitteita vertailtaessa katsotaan yleensä niiden luotettavuutta, kokoa, tehokkuutta, säätötarkkuutta ja tietysti hintaa. Tyypillisiä tehoelektroniikkalaitteita ovat taajuudenmuuttajat, UPS (Uninterruptible Power Supply) -laitteet, hitsauskoneet, induktiokuumentimet sekä erilaiset teholähteet. Perinteisesti näiden laitteiden ohjaus toteutetaan käyttäen mikroprosessoreja, ASIC- (Application Specific Integrated Circuit) tai IC (Intergrated Circuit) -piirejä sekä analogisia säätimiä. Tässä tutkimuksessa on analysoitu FPGA (Field Programmable Gate Array) -piirien soveltuvuutta tehoelektroniikan ohjaukseen. FPGA-piirien rakenne muodostuu erilaisista loogisista elementeistä ja niiden välisistä yhdysjohdoista.Loogiset elementit ovat porttipiirejä ja kiikkuja. Yhdysjohdot ja loogiset elementit ovat piirissä kiinteitä eikä koostumusta tai lukumäärää voi jälkikäteen muuttaa. Ohjelmoitavuus syntyy elementtien välisistä liitännöistä. Piirissä on lukuisia, jopa miljoonia kytkimiä, joiden asento voidaan asettaa. Siten piirin peruselementeistä voidaan muodostaa lukematon määrä erilaisia toiminnallisia kokonaisuuksia. FPGA-piirejä on pitkään käytetty kommunikointialan tuotteissa ja siksi niiden kehitys on viime vuosina ollut nopeaa. Samalla hinnat ovat pudonneet. Tästä johtuen FPGA-piiristä on tullut kiinnostava vaihtoehto myös tehoelektroniikkalaitteiden ohjaukseen. Väitöstyössä FPGA-piirien käytön soveltuvuutta on tutkittu käyttäen kahta vaativaa ja erilaista käytännön tehoelektroniikkalaitetta: taajuudenmuuttajaa ja hitsauskonetta. Molempiin testikohteisiin rakennettiin alan suomalaisten teollisuusyritysten kanssa soveltuvat prototyypit,joiden ohjauselektroniikka muutettiin FPGA-pohjaiseksi. Lisäksi kehitettiin tätä uutta tekniikkaa hyödyntävät uudentyyppiset ohjausmenetelmät. Prototyyppien toimivuutta verrattiin vastaaviin perinteisillä menetelmillä ohjattuihin kaupallisiin tuotteisiin ja havaittiin FPGA-piirien mahdollistaman rinnakkaisen laskennantuomat edut molempien tehoelektroniikkalaitteiden toimivuudessa. Työssä on myösesitetty uusia menetelmiä ja työkaluja FPGA-pohjaisen säätöjärjestelmän kehitykseen ja testaukseen. Esitetyillä menetelmillä tuotteiden kehitys saadaan mahdollisimman nopeaksi ja tehokkaaksi. Lisäksi työssä on kehitetty FPGA:n sisäinen ohjaus- ja kommunikointiväylärakenne, joka palvelee tehoelektroniikkalaitteiden ohjaussovelluksia. Uusi kommunikointirakenne edistää lisäksi jo tehtyjen osajärjestelmien uudelleen käytettävyyttä tulevissa sovelluksissa ja tuotesukupolvissa.

Three-dimensional hydrodynamic numerical modelling of fold nappe formation, basement-cover deformation and slab detachment with applications to the helvetic nappe system (W Switzerland)

Many three-dimensional (3-D) structures in rock, which formed during the deformation of the Earth's crust and lithosphere, are controlled by a difference in mechanical strength between rock units and are often the result of a geometrical instability. Such structures are, for example, folds, pinch-and-swell structures (due to necking) or cuspate-lobate structures (mullions). These struc-tures occur from the centimeter to the kilometer scale and the related deformation processes con-trol the formation of, for example, fold-and-thrust belts and extensional sedimentary basins or the deformation of the basement-cover interface. The 2-D deformation processes causing these structures are relatively well studied, however, several processes during large-strain 3-D defor-mation are still incompletely understood. One of these 3-D processes is the lateral propagation of these structures, such as fold and cusp propagation in a direction orthogonal to the shortening direction or neck propagation in direction orthogonal to the extension direction. Especially, we are interested in fold nappes which are recumbent folds with amplitudes usually exceeding 10 km and they have been presumably formed by ductile shearing. They often exhibit a constant sense of shearing and a non-linear increase of shear strain towards their overturned limb. The fold axes of the Morcles fold nappe in western Switzerland plunges to the ENE whereas the fold axes in the more eastern Doldenhorn nappe plunges to the WSW. These opposite plunge direc-tions characterize the Rawil depression (Wildstrubel depression). The Morcles nappe is mainly the result of layer parallel contraction and shearing. During the compression the massive lime-stones were more competent than the surrounding marls and shales, which led to the buckling characteristics of the Morcles nappe, especially in the north-dipping normal limb. The Dolden-horn nappe exhibits only a minor overturned fold limb. There are still no 3-D numerical studies which investigate the fundamental dynamics of the formation of the large-scale 3-D structure including the Morcles and Doldenhorn nappes and the related Rawil depression. We study the 3-D evolution of geometrical instabilities and fold nappe formation with numerical simulations based on the finite element method (FEM). Simulating geometrical instabilities caused by sharp variations of mechanical strength between rock units requires a numerical algorithm that can accurately resolve material interfaces for large differences in material properties (e.g. between limestone and shale) and for large deformations. Therefore, our FE algorithm combines a nu-merical contour-line technique and a deformable Lagrangian mesh with re-meshing. With this combined method it is possible to accurately follow the initial material contours with the FE mesh and to accurately resolve the geometrical instabilities. The algorithm can simulate 3-D de-formation for a visco-elastic rheology. The viscous rheology is described by a power-law flow law. The code is used to study the 3-D fold nappe formation, the lateral propagation of folding and also the lateral propagation of cusps due to initial half graben geometry. Thereby, the small initial geometrical perturbations for folding and necking are exactly followed by the FE mesh, whereas the initial large perturbation describing a half graben is defined by a contour line inter-secting the finite elements. Further, the 3-D algorithm is applied to 3-D viscous nacking during slab detachment. The results from various simulations are compared with 2-D resulats and a 1-D analytical solution. -- On retrouve beaucoup de structures en 3 dimensions (3-D) dans les roches qui ont pour origines une déformation de la lithosphère terrestre. Ces structures sont par exemple des plis, des boudins (pinch-and-swell) ou des mullions (cuspate-lobate) et sont présentés de l'échelle centimétrique à kilométrique. Mécaniquement, ces structures peuvent être expliquées par une différence de résistance entre les différentes unités de roches et sont généralement le fruit d'une instabilité géométrique. Ces différences mécaniques entre les unités contrôlent non seulement les types de structures rencontrées, mais également le type de déformation (thick skin, thin skin) et le style tectonique (bassin d'avant pays, chaîne d'avant pays). Les processus de la déformation en deux dimensions (2-D) formant ces structures sont relativement bien compris. Cependant, lorsque l'on ajoute la troisiéme dimension, plusieurs processus ne sont pas complètement compris lors de la déformation à large échelle. L'un de ces processus est la propagation latérale des structures, par exemple la propagation de plis ou de mullions dans la direction perpendiculaire à l'axe de com-pression, ou la propagation des zones d'amincissement des boudins perpendiculairement à la direction d'extension. Nous sommes particulièrement intéressés les nappes de plis qui sont des nappes de charriage en forme de plis couché d'une amplitude plurikilométrique et étant formées par cisaillement ductile. La plupart du temps, elles exposent un sens de cisaillement constant et une augmentation non linéaire de la déformation vers la base du flanc inverse. Un exemple connu de nappes de plis est le domaine Helvétique dans les Alpes de l'ouest. Une de ces nap-pes est la Nappe de Morcles dont l'axe de pli plonge E-NE tandis que de l'autre côté de la dépression du Rawil (ou dépression du Wildstrubel), la nappe du Doldenhorn (équivalent de la nappe de Morcles) possède un axe de pli plongeant O-SO. La forme particulière de ces nappes est due à l'alternance de couches calcaires mécaniquement résistantes et de couches mécanique-ment faibles constituées de schistes et de marnes. Ces différences mécaniques dans les couches permettent d'expliquer les plissements internes à la nappe, particulièrement dans le flanc inver-se de la nappe de Morcles. Il faut également noter que le développement du flanc inverse des nappes n'est pas le même des deux côtés de la dépression de Rawil. Ainsi la nappe de Morcles possède un important flanc inverse alors que la nappe du Doldenhorn en est presque dépour-vue. A l'heure actuelle, aucune étude numérique en 3-D n'a été menée afin de comprendre la dynamique fondamentale de la formation des nappes de Morcles et du Doldenhorn ainsi que la formation de la dépression de Rawil. Ce travail propose la première analyse de l'évolution 3-D des instabilités géométriques et de la formation des nappes de plis en utilisant des simulations numériques. Notre modèle est basé sur la méthode des éléments finis (FEM) qui permet de ré-soudre avec précision les interfaces entre deux matériaux ayant des propriétés mécaniques très différentes (par exemple entre les couches calcaires et les couches marneuses). De plus nous utilisons un maillage lagrangien déformable avec une fonction de re-meshing (production d'un nouveau maillage). Grâce à cette méthode combinée il nous est possible de suivre avec précisi-on les interfaces matérielles et de résoudre avec précision les instabilités géométriques lors de la déformation de matériaux visco-élastiques décrit par une rhéologie non linéaire (n>1). Nous uti-lisons cet algorithme afin de comprendre la formation des nappes de plis, la propagation latérale du plissement ainsi que la propagation latérale des structures de type mullions causé par une va-riation latérale de la géométrie (p.ex graben). De plus l'algorithme est utilisé pour comprendre la dynamique 3-D de l'amincissement visqueux et de la rupture de la plaque descendante en zone de subduction. Les résultats obtenus sont comparés à des modèles 2-D et à la solution analytique 1-D. -- Viele drei dimensionale (3-D) Strukturen, die in Gesteinen vorkommen und durch die Verfor-mung der Erdkruste und Litosphäre entstanden sind werden von den unterschiedlichen mechani-schen Eigenschaften der Gesteinseinheiten kontrolliert und sind häufig das Resulat von geome-trischen Istabilitäten. Zu diesen strukturen zählen zum Beispiel Falten, Pich-and-swell Struktu-ren oder sogenannte Cusbate-Lobate Strukturen (auch Mullions). Diese Strukturen kommen in verschiedenen Grössenordungen vor und können Masse von einigen Zentimeter bis zu einigen Kilometer aufweisen. Die mit der Entstehung dieser Strukturen verbundenen Prozesse kontrol-lieren die Entstehung von Gerbirgen und Sediment-Becken sowie die Verformung des Kontaktes zwischen Grundgebirge und Stedimenten. Die zwei dimensionalen (2-D) Verformungs-Prozesse die zu den genannten Strukturen führen sind bereits sehr gut untersucht. Einige Prozesse wäh-rend starker 3-D Verformung sind hingegen noch unvollständig verstanden. Einer dieser 3-D Prozesse ist die seitliche Fortpflanzung der beschriebenen Strukturen, so wie die seitliche Fort-pflanzung von Falten und Cusbate-Lobate Strukturen senkrecht zur Verkürzungsrichtung und die seitliche Fortpflanzung von Pinch-and-Swell Strukturen othogonal zur Streckungsrichtung. Insbesondere interessieren wir uns für Faltendecken, liegende Falten mit Amplituden von mehr als 10 km. Faltendecken entstehen vermutlich durch duktile Verscherung. Sie zeigen oft einen konstanten Scherungssinn und eine nicht-lineare zunahme der Scherverformung am überkipp-ten Schenkel. Die Faltenachsen der Morcles Decke in der Westschweiz fallen Richtung ONO während die Faltenachsen der östicher gelegenen Doldenhorn Decke gegen WSW einfallen. Diese entgegengesetzten Einfallrichtungen charakterisieren die Rawil Depression (Wildstrubel Depression). Die Morcles Decke ist überwiegend das Resultat von Verkürzung und Scherung parallel zu den Sedimentlagen. Während der Verkürzung verhielt sich der massive Kalkstein kompetenter als der Umliegende Mergel und Schiefer, was zur Verfaltetung Morcles Decke führ-te, vorallem in gegen Norden eifallenden überkippten Schenkel. Die Doldenhorn Decke weist dagegen einen viel kleineren überkippten Schenkel und eine stärkere Lokalisierung der Verfor-mung auf. Bis heute gibt es keine 3-D numerischen Studien, die die fundamentale Dynamik der Entstehung von grossen stark verformten 3-D Strukturen wie den Morcles und Doldenhorn Decken sowie der damit verbudenen Rawil Depression untersuchen. Wir betrachten die 3-D Ent-wicklung von geometrischen Instabilitäten sowie die Entstehung fon Faltendecken mit Hilfe von numerischen Simulationen basiert auf der Finite Elemente Methode (FEM). Die Simulation von geometrischen Instabilitäten, die aufgrund von Änderungen der Materialeigenschaften zwischen verschiedenen Gesteinseinheiten entstehen, erfortert einen numerischen Algorithmus, der in der Lage ist die Materialgrenzen mit starkem Kontrast der Materialeigenschaften (zum Beispiel zwi-schen Kalksteineinheiten und Mergel) für starke Verfomung genau aufzulösen. Um dem gerecht zu werden kombiniert unser FE Algorithmus eine numerische Contour-Linien-Technik und ein deformierbares Lagranges Netz mit Re-meshing. Mit dieser kombinierten Methode ist es mög-lich den anfänglichen Materialgrenzen mit dem FE Netz genau zu folgen und die geometrischen Instabilitäten genügend aufzulösen. Der Algorithmus ist in der Lage visko-elastische 3-D Ver-formung zu rechnen, wobei die viskose Rheologie mit Hilfe eines power-law Fliessgesetzes beschrieben wird. Mit dem numerischen Algorithmus untersuchen wir die Entstehung von 3-D Faltendecken, die seitliche Fortpflanzung der Faltung sowie der Cusbate-Lobate Strukturen die sich durch die Verkürzung eines mit Sediment gefüllten Halbgraben bilden. Dabei werden die anfänglichen geometrischen Instabilitäten der Faltung exakt mit dem FE Netz aufgelöst wäh-rend die Materialgranzen des Halbgrabens die Finiten Elemente durchschneidet. Desweiteren wird der 3-D Algorithmus auf die Einschnürung während der 3-D viskosen Plattenablösung und Subduktion angewandt. Die 3-D Resultate werden mit 2-D Ergebnissen und einer 1-D analyti-schen Lösung verglichen.