Engineering Complex Computational Ecosystems

Self-organising pervasive ecosystems of devices are set to become a major vehicle for delivering infrastructure and end-user services. The inherent complexity of such systems poses new challenges to those who want to dominate it by applying the principles of engineering. The recent growth in number and distribution of devices with decent computational and communicational abilities, that suddenly accelerated with the massive diffusion of smartphones and tablets, is delivering a world with a much higher density of devices in space. Also, communication technologies seem to be focussing on short-range device-to-device (P2P) interactions, with technologies such as Bluetooth and Near-Field Communication gaining greater adoption. Locality and situatedness become key to providing the best possible experience to users, and the classic model of a centralised, enormously powerful server gathering and processing data becomes less and less efficient with device density. Accomplishing complex global tasks without a centralised controller responsible of aggregating data, however, is a challenging task. In particular, there is a local-to-global issue that makes the application of engineering principles challenging at least: designing device-local programs that, through interaction, guarantee a certain global service level. In this thesis, we first analyse the state of the art in coordination systems, then motivate the work by describing the main issues of pre-existing tools and practices and identifying the improvements that would benefit the design of such complex software ecosystems. The contribution can be divided in three main branches. First, we introduce a novel simulation toolchain for pervasive ecosystems, designed for allowing good expressiveness still retaining high performance. Second, we leverage existing coordination models and patterns in order to create new spatial structures. Third, we introduce a novel language, based on the existing ``Field Calculus'' and integrated with the aforementioned toolchain, designed to be usable for practical aggregate programming.

Formalizing languages for service oriented computing

Service Oriented Computing is a new programming paradigm for addressing distributed system design issues. Services are autonomous computational entities which can be dynamically discovered and composed in order to form more complex systems able to achieve different kinds of task. E-government, e-business and e-science are some examples of the IT areas where Service Oriented Computing will be exploited in the next years. At present, the most credited Service Oriented Computing technology is that of Web Services, whose specifications are enriched day by day by industrial consortia without following a precise and rigorous approach. This PhD thesis aims, on the one hand, at modelling Service Oriented Computing in a formal way in order to precisely define the main concepts it is based upon and, on the other hand, at defining a new approach, called bipolar approach, for addressing system design issues by synergically exploiting choreography and orchestration languages related by means of a mathematical relation called conformance. Choreography allows us to describe systems of services from a global view point whereas orchestration supplies a means for addressing such an issue from a local perspective. In this work we present SOCK, a process algebra based language inspired by the Web Service orchestration language WS-BPEL which catches the essentials of Service Oriented Computing. From the definition of SOCK we will able to define a general model for dealing with Service Oriented Computing where services and systems of services are related to the design of finite state automata and process algebra concurrent systems, respectively. Furthermore, we introduce a formal language for dealing with choreography. Such a language is equipped with a formal semantics and it forms, together with a subset of the SOCK calculus, the bipolar framework. Finally, we present JOLIE which is a Java implentation of a subset of the SOCK calculus and it is part of the bipolar framework we intend to promote.

Supramolecular hybrid organic-inorganic multicomponent architectures in solution and on surface

Supramolecular architectures can be built-up from a single molecular component (building block) to obtain a complex of organic or inorganic interactions creating a new emergent condensed phase of matter, such as gels, liquid crystals and solid crystal. Further the generation of multicomponent supramolecular hybrid architecture, a mix of organic and inorganic components, increases the complexity of the condensed aggregate with functional properties useful for important areas of research, like material science, medicine and nanotechnology. One may design a molecule storing a recognition pattern and programming a informed self-organization process enables to grow-up into a hierarchical architecture. From a molecular level to a supramolecular level, in a bottom-up fashion, it is possible to create a new emergent structure-function, where the system, as a whole, is open to its own environment to exchange energy, matter and information. “The emergent property of the whole assembly is superior to the sum of a singles parts”. In this thesis I present new architectures and functional materials built through the selfassembly of guanosine, in the absence or in the presence of a cation, in solution and on the surface. By appropriate manipulation of intermolecular non-covalent interactions the spatial (structural) and temporal (dynamic) features of these supramolecular architectures are controlled. Guanosine G7 (5',3'-di-decanoil-deoxi-guanosine) is able to interconvert reversibly between a supramolecular polymer and a discrete octameric species by dynamic cation binding and release. Guanosine G16 (2',3'-O-Isopropylidene-5'-O-decylguanosine) shows selectivity binding from a mix of different cation's nature. Remarkably, reversibility, selectivity, adaptability and serendipity are mutual features to appreciate the creativity of a molecular self-organization complex system into a multilevelscale hierarchical growth. The creativity - in general sense, the creation of a new thing, a new thinking, a new functionality or a new structure - emerges from a contamination process of different disciplines such as biology, chemistry, physics, architecture, design, philosophy and science of complexity.

Tecniche di ottimizzazione del software per sistemi su singolo chip per applicazioni di Nomadic Computing

I moderni sistemi embedded sono equipaggiati con risorse hardware che consentono l’esecuzione di applicazioni molto complesse come il decoding audio e video. La progettazione di simili sistemi deve soddisfare due esigenze opposte. Da un lato è necessario fornire un elevato potenziale computazionale, dall’altro bisogna rispettare dei vincoli stringenti riguardo il consumo di energia. Uno dei trend più diffusi per rispondere a queste esigenze opposte è quello di integrare su uno stesso chip un numero elevato di processori caratterizzati da un design semplificato e da bassi consumi. Tuttavia, per sfruttare effettivamente il potenziale computazionale offerto da una batteria di processoriè necessario rivisitare pesantemente le metodologie di sviluppo delle applicazioni. Con l’avvento dei sistemi multi-processore su singolo chip (MPSoC) il parallel programming si è diffuso largamente anche in ambito embedded. Tuttavia, i progressi nel campo della programmazione parallela non hanno mantenuto il passo con la capacità di integrare hardware parallelo su un singolo chip. Oltre all’introduzione di multipli processori, la necessità di ridurre i consumi degli MPSoC comporta altre soluzioni architetturali che hanno l’effetto diretto di complicare lo sviluppo delle applicazioni. Il design del sottosistema di memoria, in particolare, è un problema critico. Integrare sul chip dei banchi di memoria consente dei tempi d’accesso molto brevi e dei consumi molto contenuti. Sfortunatamente, la quantità di memoria on-chip che può essere integrata in un MPSoC è molto limitata. Per questo motivo è necessario aggiungere dei banchi di memoria off-chip, che hanno una capacità molto maggiore, come maggiori sono i consumi e i tempi d’accesso. La maggior parte degli MPSoC attualmente in commercio destina una parte del budget di area all’implementazione di memorie cache e/o scratchpad. Le scratchpad (SPM) sono spesso preferite alle cache nei sistemi MPSoC embedded, per motivi di maggiore predicibilità, minore occupazione d’area e – soprattutto – minori consumi. Per contro, mentre l’uso delle cache è completamente trasparente al programmatore, le SPM devono essere esplicitamente gestite dall’applicazione. Esporre l’organizzazione della gerarchia di memoria ll’applicazione consente di sfruttarne in maniera efficiente i vantaggi (ridotti tempi d’accesso e consumi). Per contro, per ottenere questi benefici è necessario scrivere le applicazioni in maniera tale che i dati vengano partizionati e allocati sulle varie memorie in maniera opportuna. L’onere di questo compito complesso ricade ovviamente sul programmatore. Questo scenario descrive bene l’esigenza di modelli di programmazione e strumenti di supporto che semplifichino lo sviluppo di applicazioni parallele. In questa tesi viene presentato un framework per lo sviluppo di software per MPSoC embedded basato su OpenMP. OpenMP è uno standard di fatto per la programmazione di multiprocessori con memoria shared, caratterizzato da un semplice approccio alla parallelizzazione tramite annotazioni (direttive per il compilatore). La sua interfaccia di programmazione consente di esprimere in maniera naturale e molto efficiente il parallelismo a livello di loop, molto diffuso tra le applicazioni embedded di tipo signal processing e multimedia. OpenMP costituisce un ottimo punto di partenza per la definizione di un modello di programmazione per MPSoC, soprattutto per la sua semplicità d’uso. D’altra parte, per sfruttare in maniera efficiente il potenziale computazionale di un MPSoC è necessario rivisitare profondamente l’implementazione del supporto OpenMP sia nel compilatore che nell’ambiente di supporto a runtime. Tutti i costrutti per gestire il parallelismo, la suddivisione del lavoro e la sincronizzazione inter-processore comportano un costo in termini di overhead che deve essere minimizzato per non comprometterre i vantaggi della parallelizzazione. Questo può essere ottenuto soltanto tramite una accurata analisi delle caratteristiche hardware e l’individuazione dei potenziali colli di bottiglia nell’architettura. Una implementazione del task management, della sincronizzazione a barriera e della condivisione dei dati che sfrutti efficientemente le risorse hardware consente di ottenere elevate performance e scalabilità. La condivisione dei dati, nel modello OpenMP, merita particolare attenzione. In un modello a memoria condivisa le strutture dati (array, matrici) accedute dal programma sono fisicamente allocate su una unica risorsa di memoria raggiungibile da tutti i processori. Al crescere del numero di processori in un sistema, l’accesso concorrente ad una singola risorsa di memoria costituisce un evidente collo di bottiglia. Per alleviare la pressione sulle memorie e sul sistema di connessione vengono da noi studiate e proposte delle tecniche di partizionamento delle strutture dati. Queste tecniche richiedono che una singola entità di tipo array venga trattata nel programma come l’insieme di tanti sotto-array, ciascuno dei quali può essere fisicamente allocato su una risorsa di memoria differente. Dal punto di vista del programma, indirizzare un array partizionato richiede che ad ogni accesso vengano eseguite delle istruzioni per ri-calcolare l’indirizzo fisico di destinazione. Questo è chiaramente un compito lungo, complesso e soggetto ad errori. Per questo motivo, le nostre tecniche di partizionamento sono state integrate nella l’interfaccia di programmazione di OpenMP, che è stata significativamente estesa. Specificamente, delle nuove direttive e clausole consentono al programmatore di annotare i dati di tipo array che si vuole partizionare e allocare in maniera distribuita sulla gerarchia di memoria. Sono stati inoltre sviluppati degli strumenti di supporto che consentono di raccogliere informazioni di profiling sul pattern di accesso agli array. Queste informazioni vengono sfruttate dal nostro compilatore per allocare le partizioni sulle varie risorse di memoria rispettando una relazione di affinità tra il task e i dati. Più precisamente, i passi di allocazione nel nostro compilatore assegnano una determinata partizione alla memoria scratchpad locale al processore che ospita il task che effettua il numero maggiore di accessi alla stessa.

Dynamic power management: from portable devices to high performance computing

Electronic applications are nowadays converging under the umbrella of the cloud computing vision. The future ecosystem of information and communication technology is going to integrate clouds of portable clients and embedded devices exchanging information, through the internet layer, with processing clusters of servers, data-centers and high performance computing systems. Even thus the whole society is waiting to embrace this revolution, there is a backside of the story. Portable devices require battery to work far from the power plugs and their storage capacity does not scale as the increasing power requirement does. At the other end processing clusters, such as data-centers and server farms, are build upon the integration of thousands multiprocessors. For each of them during the last decade the technology scaling has produced a dramatic increase in power density with significant spatial and temporal variability. This leads to power and temperature hot-spots, which may cause non-uniform ageing and accelerated chip failure. Nonetheless all the heat removed from the silicon translates in high cooling costs. Moreover trend in ICT carbon footprint shows that run-time power consumption of the all spectrum of devices accounts for a significant slice of entire world carbon emissions. This thesis work embrace the full ICT ecosystem and dynamic power consumption concerns by describing a set of new and promising system levels resource management techniques to reduce the power consumption and related issues for two corner cases: Mobile Devices and High Performance Computing.

Interoperability Challenges in Internet of Things Systems: a Service-Oriented Computing Approach

Internet of Things systems are pervasive systems evolved from cyber-physical to large-scale systems. Due to the number of technologies involved, software development involves several integration challenges. Among them, the ones preventing proper integration are those related to the system heterogeneity, and thus addressing interoperability issues. From a software engineering perspective, developers mostly experience the lack of interoperability in the two phases of software development: programming and deployment. On the one hand, modern software tends to be distributed in several components, each adopting its most-appropriate technology stack, pushing programmers to code in a protocol- and data-agnostic way. On the other hand, each software component should run in the most appropriate execution environment and, as a result, system architects strive to automate the deployment in distributed infrastructures. This dissertation aims to improve the development process by introducing proper tools to handle certain aspects of the system heterogeneity. Our effort focuses on three of these aspects and, for each one of those, we propose a tool addressing the underlying challenge. The first tool aims to handle heterogeneity at the transport and application protocol level, the second to manage different data formats, while the third to obtain optimal deployment. To realize the tools, we adopted a linguistic approach, i.e.\ we provided specific linguistic abstractions that help developers to increase the expressive power of the programming language they use, writing better solutions in more straightforward ways. To validate the approach, we implemented use cases to show that the tools can be used in practice and that they help to achieve the expected level of interoperability. In conclusion, to move a step towards the realization of an integrated Internet of Things ecosystem, we target programmers and architects and propose them to use the presented tools to ease the software development process.

Constraint Programming-based Job Dispatching for Modern HPC Applications

A High-Performance Computing job dispatcher is a critical software that assigns the finite computing resources to submitted jobs. This resource assignment over time is known as the on-line job dispatching problem in HPC systems. The fact the problem is on-line means that solutions must be computed in real-time, and their required time cannot exceed some threshold to do not affect the normal system functioning. In addition, a job dispatcher must deal with a lot of uncertainty: submission times, the number of requested resources, and duration of jobs. Heuristic-based techniques have been broadly used in HPC systems, at the cost of achieving (sub-)optimal solutions in a short time. However, the scheduling and resource allocation components are separated, thus generates a decoupled decision that may cause a performance loss. Optimization-based techniques are less used for this problem, although they can significantly improve the performance of HPC systems at the expense of higher computation time. Nowadays, HPC systems are being used for modern applications, such as big data analytics and predictive model building, that employ, in general, many short jobs. However, this information is unknown at dispatching time, and job dispatchers need to process large numbers of them quickly while ensuring high Quality-of-Service (QoS) levels. Constraint Programming (CP) has been shown to be an effective approach to tackle job dispatching problems. However, state-of-the-art CP-based job dispatchers are unable to satisfy the challenges of on-line dispatching, such as generate dispatching decisions in a brief period and integrate current and past information of the housing system. Given the previous reasons, we propose CP-based dispatchers that are more suitable for HPC systems running modern applications, generating on-line dispatching decisions in a proper time and are able to make effective use of job duration predictions to improve QoS levels, especially for workloads dominated by short jobs.

Monitoring and Prediction of Thermal Emergencies in High Performance Computing Systems

Modern scientific discoveries are driven by an unsatisfiable demand for computational resources. High-Performance Computing (HPC) systems are an aggregation of computing power to deliver considerably higher performance than one typical desktop computer can provide, to solve large problems in science, engineering, or business. An HPC room in the datacenter is a complex controlled environment that hosts thousands of computing nodes that consume electrical power in the range of megawatts, which gets completely transformed into heat. Although a datacenter contains sophisticated cooling systems, our studies indicate quantitative evidence of thermal bottlenecks in real-life production workload, showing the presence of significant spatial and temporal thermal and power heterogeneity. Therefore minor thermal issues/anomalies can potentially start a chain of events that leads to an unbalance between the amount of heat generated by the computing nodes and the heat removed by the cooling system originating thermal hazards. Although thermal anomalies are rare events, anomaly detection/prediction in time is vital to avoid IT and facility equipment damage and outage of the datacenter, with severe societal and business losses. For this reason, automated approaches to detect thermal anomalies in datacenters have considerable potential. This thesis analyzed and characterized the power and thermal characteristics of a Tier0 datacenter (CINECA) during production and under abnormal thermal conditions. Then, a Deep Learning (DL)-powered thermal hazard prediction framework is proposed. The proposed models are validated against real thermal hazard events reported for the studied HPC cluster while in production. This thesis is the first empirical study of thermal anomaly detection and prediction techniques of a real large-scale HPC system to the best of my knowledge. For this thesis, I used a large-scale dataset, monitoring data of tens of thousands of sensors for around 24 months with a data collection rate of around 20 seconds.

PCM-based in-memory computing: architectures, circuits and applications

Analog In-memory Computing (AIMC) has been proposed in the context of Beyond Von Neumann architectures as a valid strategy to reduce internal data transfers energy consumption and latency, and to improve compute efficiency. The aim of AIMC is to perform computations within the memory unit, typically leveraging the physical features of memory devices. Among resistive Non-volatile Memories (NVMs), Phase-change Memory (PCM) has become a promising technology due to its intrinsic capability to store multilevel data. Hence, PCM technology is currently investigated to enhance the possibilities and the applications of AIMC. This thesis aims at exploring the potential of new PCM-based architectures as in-memory computational accelerators. In a first step, a preliminar experimental characterization of PCM devices has been carried out in an AIMC perspective. PCM cells non-idealities, such as time-drift, noise, and non-linearity have been studied to develop a dedicated multilevel programming algorithm. Measurement-based simulations have been then employed to evaluate the feasibility of PCM-based operations in the fields of Deep Neural Networks (DNNs) and Structural Health Monitoring (SHM). Moreover, a first testchip has been designed and tested to evaluate the hardware implementation of Multiply-and-Accumulate (MAC) operations employing PCM cells. This prototype experimentally demonstrates the possibility to reach a 95% MAC accuracy with a circuit-level compensation of cells time drift and non-linearity. Finally, empirical circuit behavior models have been included in simulations to assess the use of this technology in specific DNN applications, and to enhance the potentiality of this innovative computation approach.