Dynamic partial reconfiguration of FPGAs allows systems to autonomously alter sections of their design during runtime based on the state of the system. This functionality provides size, weight, and power benefits that are useful in extreme environments such as space. Therefore, NASA has requested research into the feasibility of using a commercial off-the-shelf software flow to convert a static HDL design to support partial reconfiguration. This project presents an analysis of this conversion process using the Xilinx Partial Reconfiguration Flow to convert the static design for the ITU G.729 Voice Decoder. This paper explores the design modifications that must be made to allow for partial reconfiguration. Furthermore, an in-depth description of how to set up the hardware platform to support the HDL application is provided. Finally, timing and size data are presented and analyzed to empirically show the benefits and limitations of using dynamic partial reconfiguration.

Modern Field-Programmable Gate Arrays (FPGAs) are no longer used to implement small “glue logic” circuitries. The high-density of reconfigurable logic resources in today’s FPGAs enable the implementation of large systems in a single chip. FPGAs are highly flexible devices; their functionality can be altered by simply loading a new binary file in their configuration memory. While the flexibility of FPGAs is comparable to General-Purpose Processors (GPPs), in the sense that different functions can be performed using the same hardware, the performance gain that can be achieved using FPGAs can be orders of magnitudes higher as FPGAs offer the ability for customisation of parallel computational architectures. Dynamic Partial Reconfiguration (DPR) allows for changing the functionality of certain blocks on the chip while the rest of the FPGA is operational. DPR has sparked the interest of researchers to explore new computational platforms where computational tasks are off-loaded from a main CPU to be executed using dedicated reconfigurable hardware accelerators configured on demand at run-time. By having a battery of custom accelerators which can be swapped in and out of the FPGA at runtime, a higher computational density can be achieved compared to static systems where the accelerators are bound to fixed locations within the chip. Furthermore, the ability of relocating these accelerators across several locations on the chip allows for the implementation of adaptive systems which can mitigate emerging faults in the FPGA chip when operating in harsh environments. By porting the appropriate fault mitigation techniques in such computational platforms, the advantages of FPGAs can be harnessed in different applications in space and military electronics where FPGAs are usually seen as unreliable devices due to their sensitivity to radiation and extreme environmental conditions. In light of the above, this thesis investigates the deployment of DPR as: 1) a method for enhancing performance by efficient exploitation of the FPGA resources, and 2) a method for enhancing the reliability of systems intended to operate in harsh environments. Achieving optimal performance in such systems requires an efficient internal configuration management system to manage the reconfiguration and execution of the reconfigurable modules in the FPGA. In addition, the system needs to support “fault-resilience” features by integrating parameterisable fault detection and recovery capabilities to meet the reliability standard of fault-tolerant applications. This thesis addresses all the design and implementation aspects of an Internal Configuration Manger (ICM) which supports a novel bitstream relocation model to enable the placement of relocatable accelerators across several locations on the FPGA chip. In addition to supporting all the configuration capabilities required to implement a Reconfigurable Operating System (ROS), the proposed ICM also supports the novel multiple-clone configuration technique which allows for cloning several instances of the same hardware accelerator at the same time resulting in much shorter configuration time compared to traditional configuration techniques. A faulttolerant (FT) version of the proposed ICM which supports a comprehensive faultrecovery scheme is also introduced in this thesis. The proposed FT-ICM is designed with a much smaller area footprint compared to Triple Modular Redundancy (TMR) hardening techniques while keeping a comparable level of fault-resilience. The capabilities of the proposed ICM system are demonstrated with two novel applications. The first application demonstrates a proof-of-concept reliable FPGA server solution used for executing encryption/decryption queries. The proposed server deploys bitstream relocation and modular redundancy to mitigate both permanent and transient faults in the device. It also deploys a novel Built-In Self- Test (BIST) diagnosis scheme, specifically designed to detect emerging permanent faults in the system at run-time. The second application is a data mining application where DPR is used to increase the computational density of a system used to implement the Frequent Itemset Mining (FIM) problem.

This project investigates a self-reconfiguring rSoC (reconfigurable System on Chip) system which automatically and dynamically loads peripheral controllers, based on the peripherals connected to the system. The Xilinx Virtex-II FGPA, which supports dynamic partial reconfiguration, is used as the experimental target. To implement the system, three main areas are investigated: the peripheral auto detection, the dynamic partial reconfiguration mechanism on the FPGA, and the supporting software. The system core is designed as two defined areas on a single FPGA chip. A fixed area is used for the constant logic circuits (such as soft-core CPU) and partial reconfiguration (PR) slots are used for changeable peripheral controllers. The autoconfiguration process involves three different steps: peripheral auto detection, loading of a peripheral hardware interface configuration, and loading of a peripheral software driver. In our system, we successfully implement the mechanism of peripheral dynamic loading on the rSoC system. Four novel features are provided in the system: 1) Peripheral auto detection. Peripheral boards are automatically detected by the system when connected to the system. 2) Peripheral controller hardware bitstream and software driver dynamic loading. The required peripheral controller hardware bitstream for the connected peripheral board is automatically searched for and loaded by the operating system, as well as the required software driver. Manual operations on these processes are also supported. 3) Individual interface to external environment. Each PR slot provides individual interface to peripheral boards. It is configured by each peripheral controller for board-specific connection. 4) The existing system is extensible. The partial reconfiguration mechanism provided in this project supports at least two PR slots. On higher capacity FPGAs, the number of PR slots could be increased. In our existing system, the time used for the dynamic partial reconfiguration process, including the hardware bitstream loading and the software driver loading, is in the order of 10-20ms, which is an insignificant fraction of the Linux boot time.

For applications such as live video processing, there is a high demand for high performance and low latency solutions. The configurable logic in FPGAs allows for custom hardware to be tailored to a specific video application. These FPGA designs require technical expertise and lengthy implementation times by vendor tools for each unique solution. This thesis presents a dynamically configurable topology as an FPGA overlay to deploy custom hardware processing pipelines during run-time by utilizing dynamic partial reconfiguration. Within the FPGA overlay, a configurable topology with a routable switch allows video streams to be copied and mixed to create complex data paths. This work demonstrates a dynamic video processing pipeline with 11 reconfigurable regions and 16 unique processing cores, allowing for billions of custom run-time configurations.

Nos últimos anos, houve um grande avanço na computação reconfigurável, em particular em hardware que emprega Field-Programmable Gate Arrays. Porém, esse aumento de capacidade e desempenho aumentou a distância entre a capacidade de projeto e a disponibilidade de tecnologia para o desenvolvimento do projeto. As linguagens de programação imperativas de alto nível, como C, são mais apropriadas para o desenvolvimento de aplicativos complexos que as linguagens de descrição de hardware. Por isso, surgiram diversas ferramentas para o desenvolvimento de hardware a partir de código em C. A ferramenta ChipCflow, da qual faz parte este projeto, é uma delas. A execução dos programas por meio dessa ferramenta será completamente baseada em seu fluxo de dados, seguindo o modelo dinâmico encontrado nas arquiteturas de computadores a fluxo de dados, aproveitando ao máximo o paralelismo considerado natural desse modelo e as características do hardware parcialmente reconfigurável. Neste projeto em particular, o objetivo é a prova de conceito (proof of concept) para a criação de instâncias, em forma de operadores, de um algoritmo ChipCflow em hardware parcialmente reconfigurável, tendo como base a plataforma Virtex da Xilinx / In recent years, reconfigurable computing has become increasingly more advanced, especially in hardware that uses Field-Programmable Gate Arrays. However, the increase of performance in FPGAs accumulated the gap between design capacity and technology for the development of the design. Imperative high-level programming languages such as C are more appropriate for the development of complex algorithms than hardware description languages (HDL). For this reason, many ANSI C-like programming tools for the development of hardware came to existence. The ChipCflow project, of which this project is part, is one of these tools. The execution of algorithms through this tool will be completely directed by data flow, according to the dynamic model found on Dataflow Architectures, taking advantage of its natural high levels of parallelism and the characteristics of the partially reconfigurable hardware. In this project, the objective is a proof of concept for the creation of instances, in the form of operators, of a ChipCflow algorithm on a partially reconfigurable hardware, taking as reference the Xilinx Virtex boards

Nos últimos anos, houve um grande avanço na computação reconfigurável, em particular em hardware que emprega Field-Programmable Gate Arrays. Porém, esse aumento de capacidade e desempenho aumentou a distância entre a capacidade de projeto e a disponibilidade de tecnologia para o desenvolvimento do projeto. As linguagens de programação imperativas de alto nível, como C, são mais apropriadas para o desenvolvimento de aplicativos complexos que as linguagens de descrição de hardware. Por isso, surgiram diversas ferramentas para o desenvolvimento de hardware a partir de código em C. A ferramenta ChipCflow, da qual faz parte este projeto, é uma delas. A execução dos programas por meio dessa ferramenta será completamente baseada em seu fluxo de dados, seguindo o modelo dinâmico encontrado nas arquiteturas de computadores a fluxo de dados, aproveitando ao máximo o paralelismo considerado natural desse modelo e as características do hardware parcialmente reconfigurável. Neste projeto em particular, o objetivo é a prova de conceito (proof of concept) para a criação de instâncias, em forma de operadores, de um algoritmo ChipCflow em hardware parcialmente reconfigurável, tendo como base a plataforma Virtex da Xilinx / In recent years, reconfigurable computing has become increasingly more advanced, especially in hardware that uses Field-Programmable Gate Arrays. However, the increase of performance in FPGAs accumulated the gap between design capacity and technology for the development of the design. Imperative high-level programming languages such as C are more appropriate for the development of complex algorithms than hardware description languages (HDL). For this reason, many ANSI C-like programming tools for the development of hardware came to existence. The ChipCflow project, of which this project is part, is one of these tools. The execution of algorithms through this tool will be completely directed by data flow, according to the dynamic model found on Dataflow Architectures, taking advantage of its natural high levels of parallelism and the characteristics of the partially reconfigurable hardware. In this project, the objective is a proof of concept for the creation of instances, in the form of operators, of a ChipCflow algorithm on a partially reconfigurable hardware, taking as reference the Xilinx Virtex boards

Video coding is widely used in our daily life. Due to its high computational complexity, hardware implementation is usually preferred. In this research, we investigate both ASIC hardware design approach and reconfigurable hardware design approach for video coding applications. First, we present a unified architecture that can perform Discrete Cosine Transform (DCT), Inverse Discrete Cosine Transform (IDCT), DCT domain motion estimation and compensation (DCT-ME/MC). Our proposed architecture is a Wavefront Array-based Processor with a highly modular structure consisting of 8*8 Processing Elements (PEs). By utilizing statistical properties and arithmetic operations, it can be used as a high performance hardware accelerator for video transcoding applications. We show how different core algorithms can be mapped onto the same hardware fabric and can be executed through the pre-defined PEs. In addition to the simplified design process of the proposed architecture and savings of the hardware resources, we also demonstrate that high throughput rate can be achieved for IDCT and DCT-MC by fully utilizing the sparseness property of DCT coefficient matrix. Compared to fixed hardware architecture using ASIC design approach, reconfigurable hardware design approach has higher flexibility, lower cost, and faster time-to-market. We propose a self-reconfigurable platform which can reconfigure the architecture of DCT computations during run-time using dynamic partial reconfiguration. The scalable architecture for DCT computations can compute different number of DCT coefficients in the zig-zag scan order to adapt to different requirements, such as power consumption, hardware resource, and performance. We propose a configuration manager which is implemented in the embedded processor in order to adaptively control the reconfiguration of scalable DCT architecture during run-time. In addition, we use LZSS algorithm for compression of the partial bitstreams and on-chip BlockRAM as a cache to reduce latency overhead for loading the partial bitstreams from the off-chip memory for run-time reconfiguration. A hardware module is designed for parallel reconfiguration of the partial bitstreams. The experimental results show that our approach can reduce the external memory accesses by 69% and can achieve 400 MBytes/s reconfiguration rate. Detailed trade-offs of power, throughput, and quality are investigated, and used as a criterion for self-reconfiguration. Prediction algorithm of zero quantized DCT (ZQDCT) to control the run-time reconfiguration of the proposed scalable architecture has been used, and 12 different modes of DCT computations including zonal coding, multi-block processing, and parallel-sequential stage modes are supported to reduce power consumptions, required hardware resources, and computation time with a small quality degradation. Detailed trade-offs of power, throughput, and quality are investigated, and used as a criterion for self-reconfiguration to meet the requirements set by the users.

Twenty-first century Field-Programmable Gate Arrays (FPGAs) are no longer used for implementing simple “glue logic” functions. They have become complex arrays of reconfigurable logic resources and memories as well as highly optimised functional blocks, capable of implementing large systems on a single chip. Moreover, Dynamic Partial Reconfiguration (DPR) capability permits to adjust some logic resources on the chip at runtime, whilst the rest are still performing active computations. During the last few years, DPR has become a hot research topic with the objective of building more reliable, efficient and powerful electronic systems. For instance, DPR can be used to mitigate spontaneously occurring bit upsets provoked by radiation, or to jiggle around the FPGA resources which progressively get damaged as the silicon ages. Moreover, DPR is the enabling technology for a new computing paradigm which combines computation in time and space. In Reconfigurable Computing (RC), a battery of computation-specific circuits (“hardware tasks”) are swapped in and out of the FPGA on demand to hold a continuous stream of input operands, computation and output results. Multitasking, adaptation and specialisation are key properties in RC, as multiple swappable tasks can run concurrently at different positions on chip, each with custom data-paths for efficient execution of specific computations. As a result, considerable computational throughput can be achieved even at low clock frequencies. However, DPR penetration in the commercial market is still testimonial, mainly due to the lack of suitable high-level design tools to exploit this technology. Indeed, currently, special skills are required to successfully develop a dynamically reconfigurable application. In light of the above, this thesis aims at bridging the gap between high-level application and low-level DPR technology. Its main objective is to develop Operating System (OS)-like support for high-level software-centric application developers in order to exploit the benefits brought about by DPR technology, without having to deal with the complex low-level hardware details. The developed solution in this thesis is named as R3TOS, which stands for Reliable Reconfigurable Real-Time Operating System. R3TOS defines a flexible infrastructure for reliably executing reconfigurable hardware-based applications under real-time constraints. In R3TOS, the hardware tasks are scheduled in order to meet their computation deadlines and allocated to non-damaged resources, keeping the system fault-free at all times. In addition, R3TOS envisages a computing framework whereby both hardware and software tasks coexist in a seamless manner, allowing the user to access the advanced computation capabilities of modern reconfigurable hardware from a software “look and feel” environment. This thesis covers all of the design and implementation aspects of R3TOS. The thesis proposes a novel EDF-based scheduling algorithm, two novel task allocation heuristics (EAC and EVC) and a novel task allocation strategy (called Snake), addressing many RC-related particularities as well as technological constraints imposed by current FPGA technology. Empirical results show that these approaches improve on the state of the art. Besides, the thesis describes a novel way to harness the internal reconfiguration mechanism of modern FPGAs to performinter-task communications and synchronisation regardless of the physical location of tasks on-chip. This paves the way for implementing more sophisticated RC solutions which were only possible in theory in the past. The thesis illustrates R3TOS through a proof-of-concept prototype with two demonstrator applications: (1) dependability oriented control of the power chain of a railway traction vehicle, and (2) datastreaming oriented Software Defined Radio (SDR).

Ces dernières années, les fournisseurs de cloud et les centres de données ont intégrés les FPGA dans leur environnement à des fins d'accélération. Cela est dû au fait que les accélérateurs à base de FPGA sont connus pour leur faible puissance et leurs bonnes performances par watt. En outre, l'introduction de la capacité de reconfiguration partielle dynamique (DPR) de certains FPGA incite les chercheurs de l'industrie et des universitaires à proposer des services de cloud FPGA virtualisés baser sur DPR. Dans la plupart des travaux existants, l'interconnexion entre les vFPGA repose soit sur les réseaux BUS ou OpenFlow. Cependant le bus et OpenFlow ne sont pas des solutions optimales pour la virtualisation.Dans cette thèse, nous avons proposé un NoC évolutif à l'exécution pour les accélérateurs basés sur FPGA virtualisés dans un cloud computing. Les composants NoC s'adapteront dynamiquement aux nombres d'accélérateurs virtualisés actifs en ajoutant et en supprimant des sous-noC. Pour minimiser la complexité de la conception de l'architecture NoC à un niveau inférieur (implémentation HDL), nous avons proposé un langage de modélisation unifié de haut niveau (UML) basé sur une ingénierie dirigée par les modèles. Une approche basée sur UML / MARTE et IP-XACT est utilisée pour définir les composants de la topologie NoC de haut niveau et générer les fichiers HDL requis. Les résultats des expériences montrent que le NoC évolutif à l'exécution peut réduire la consommation d'énergie de 17%. La caractérisation NoC sur la modélisation de haut niveau basée sur MDE réduit également le temps de conception de 25%. / In the last few years, cloud providers and data centers have been integrating FPGAs in their environment for acceleration purpose. This is due to the fact that FPGA based accelerator are known for their lower power and good performance per watt. Moreover, the introduction of the ability for dynamic partial reconfiguration (DPR) of some FPGAs trigger researchers in both industry and academics to propose DPR based virtualized FPGA (vFPGA) cloud services. In most of the existing works, the interconnection between the vFPGAs relies either on BUS or OpenFlow networks. However, both the bus and OpenFlow are not virtualization-aware and optimal solutions. In this thesis, we have proposed a virtualization-aware dynamically scalable NoC for virtualized FPGA accelerators in cloud computing. The NoC components will adapt to the number of active virtualized accelerator dynamically by adding and removing sub-NoCs. To minimize the complexity of NoC architecture design at a low level (HDL implementation), we have proposed a Model-Driven Engineering (MDE) based high-level unified modeling language (UML). A UML/MARTE and IP-XACT based approach are used to define the NoC Topology components at a high-level and generate the required HDL files. Experiment results show that the dynamically scalable NoC can reduce the power consumption by 17%. The MDE based high-level modeling based NoC characterization also reduce the design time by 25%.

Les systèmes adaptatifs basés sur les architectures FPGA (Field-Programmable Gate Arrays) peuvent bénéficier grandement de la grande flexibilité offerte par la reconfiguration partielle dynamique (DPR). Grâce au DPR, les tâches matérielles composant un système adaptatif peuvent être allouées et re-allouées à la demande ou en fonction de l'environnement dynamique. Les flots de conceptions disponibles et les outils commerciaux ont évolué pour répondre aux exigences des architectures reconfigurables qui sont toutefois limitées dans leurs fonctionnalités. Ces outils ne permettent pas un placement et une relocation efficaces de tâches matérielles de tailles variables. L'objectif principal de ces travaux de thèse consiste à proposer des nouvelles méthodologies et de nouvelles approches pour faciliter au concepteur la phase de conception d'un système adaptatif reconfigurable opérationnelle, valide, optimisé et adapté aux changements dynamiques de l'environnement. La première contribution de cette thèse porte sur la problématique de la relocation des tâches matérielles de tailles différentes. Une méthodologie de conception est proposée pour répondre à un problème majeur des mécanismes de relogement : le stockage d'une unique bitstream de configuration pour réduire les besoins de la mémoire et pour accroître la réutilisable des modules matériels générés. Une technique de partitionnement de la région reconfigurable est appliquée dans la méthodologie de relogement proposée pour augmenter l'efficacité d'utilisation des ressources matérielles dans le cas des tâches reconfigurables de tailles variables. Cette méthodologie prend en compte aussi la communication entre différentes régions reconfigurables et la région statique. Pour valider la méthode, plusieurs études de cas sont implémentées. Cette validation montre une utilisation efficace des ressources matérielles ainsi une réduction importante du temps de reconfiguration. La deuxième partie de cette thèse présente et détaille une formulation mathématique afin d'automatiser le floorplanning des zones reconfigurables dans les FPGAs. Les algorithmes de recherche présentés dans cette thèse sont basés sur la technique d'optimisation PLMNE (programmation linéaire mixte en nombres entiers). Ces algorithmes permettent de définir automatiquement l'emplacement, la taille et la forme de la zone reconfigurable dynamique. Nous nous intéressons principalement dans cette recherche à la satisfaction des contraintes de placement des zones reconfigurables et celles liées à la relocation. De plus, nous considérons l’optimisation des ressources matérielles dans le FPGA en tenant compte des tâches de tailles variables. Finalement, une évaluation de l'approche proposée est présentée / Adaptive systems based on Field-Programmable Gate Arrays (FPGA) architectures can benefit greatly from the high degree of flexibility offered by dynamic partial reconfiguration (DPR). Thanks to DPR, hardware tasks composing an adaptive system can be allocated and relocated on demand or depending on the dynamically changing environment. Existing design flows and commercial tools have evolved to meet the requirements of reconfigurables architectures, but that are limited in functionality. These tools do not allow an efficient placement and relocation of variable-sized hardware tasks. The main objective of this thesis is to propose a new methodology and a new approaches to facilitate to the designers the design phase of an adaptive and reconfigurable system and to make it operational, valid, optimized and adapted to dynamic changes in the environment. The first contribution of this thesis deals with the issues of relocation of variable-sized hardware tasks. A design methodology is proposed to address a major problem of relocation mechanisms: storing a single configuration bitstream to reduce memory requirements and increasing the reusability of generating hardware modules. A reconfigurable region partitioning technique is applied in this proposed relocation methodology to increase the efficiency of use of hardware resources in the case of reconfigurable tasks of variable sizes. This methodology also takes into account communication between different reconfigurable regions and the static region. To validate the design method, several cases studies are implemented. This validation shows an efficient use of hardware resources and a significant reduction in reconfiguration time. The second part of this thesis presents and details a mathematical formulations in order to automate the floorplanning of the reconfigurable regions in the FPGAs. The algorithms presented in this thesis are based on the optimization technique MILP (mixed integer linear programming). These algorithms allow to define automatically the location, the size and the shape of the dynamic reconfigurable region. We are mainly interested in this research to satisfy the constraints of placement of the reconfigurable zones and those related to the relocation. In addition, we consider the optimization of the hardware resources in the FPGA taking into account the tasks of variable sizes. Finally, an evaluation of the proposed approach is presented