Vector graphics provide smooth, resolution-independent images and are used for user interfaces, illustrations,fonts and more in a wide range of applications.During the last years, handheld devices have become increasingly powerful and feature-rich. It is expectedthat an increasing number of devices will contain dedicated GPUs (graphics processing units)capable of high quality 3d graphics for games. It is of interest to use the same hardware for acceleratingvector graphics.OpenVG is a new API for vector graphics rendering on a wide range of devices from desktop to handheld.Implementations can use different algorithms and ways of accelerating the rendering process in hardware,transparent from the user application.State of the art vector graphics solutions perform much processing in the CPU, transfer large amounts ofvertex and polygon data from the CPU to GPU, and generally use the GPU in a suboptimal way. Moreefficient approaches are desirable.Recently developed algorithms provide efficient curve rendering with little CPU overhead and a significantreduction in vertex and polygon count. Some issues remain before the approach can be used forrendering in an OpenVG implementation.This thesis builds on these algorithms to develop an approach that can be used for a conformant OpenVGimplementation. A number of issues, mainly related to precision, robustness and missing features, areidentified. Solutions are suggested and either implemented in a prototype or left as future work.Preliminary tests compare the new approach to traditional approximation with line segments.Vertex and triangle count as well as the simulated tile list counts are lowered significantly and CPUoverhead from subdivision is avoided or greatly reduced in many common cases. CPU overhead fromtessellation is eliminated through the use of an improved stencil buffer technique.Data-sets with different properties show varying amounts of improvement from the new approach. Forsome data-sets, vertex and triangle count is lowered by up to 70% and subdivision is completely avoided,while for others there is no improvement.

In this paper, we present our implementation of an Auto tuning system, written in C++, which incorporate the use of OpenCL kernels. We deploy this approach on different GPU architectures, evaluating the performance of the approach. Our main focus is to easily generate tuned code, that would otherwise require a large amount of empirical testing, and then run it on any kind of device. This is achieved through the auto tuning framework, which will create different kernels, compile and run them on the device and output the best performing kernel on the given platform.BLAS is much used in performance critical applications, and is a good candidate for execution on GPUs due to its potential performance increase. Our implementation was benchmarked on various of test environments, with different GPUs, where we achieved comparable results to the ViennaCL library. We also tested against the native vendor specific BLAS libraries from AMD and NVIDIA.

A dream about evolvable structures that change to fit its environment could be a peak into the future.Cellular automata (CA) being a simple discrete model, it has the ability to simulate biology by growing, reproducing and dying. Along with genetic algorithms, they both simulates biological systems that can be used to realize this dream.In this thesis, a skyscraper is grown using multiple cellular automata. The skyscraper is grown in a CA simulator and visualizer made for this thesis. The result is a stable structure containing floors, walls, windows and ceilings with lights.Genetic algorithms have been used to grow electrical wiring from a power source in the basement up to power outlets on each floor, powering the lights.The dream is a house that covers all your needs.This thesis is a proof of concept, that it is possible to grow a stable skyscraper using a CA with multiple sub-CAs growing lights and electrical wiring inside.The project is in the area of unconventional computation, done at NTNU Trondheim.

Med sin store vekst, har internett utviklet seg til et lukrativt domene for organisert kriminalitet. Som andre typer organisert kriminalitet er mesteparten av aktiviteten motivert av økonomisk gevinst. I tillegg til økonomisk motiverte trusselen aktører er noen tilsynelatende drevet av politiske motiver, som nasjonalstaters etterretningsorganisasjoner og cyberterrorister. En viktig del av datakriminalitet er å bryte seg inn i datasystemer og sikre fremtidig kontroll over systemene. Når nettkriminelle har klart å få tilgang til et system installerer de ofte et skjult program for å sikre fremtidige tilgang. Dette programmet kalles en bot og rekrutterer den kompromitterte maskinen inn i et botnet. Flere boter under en felles sentral administrasjon kalles et botnet. Denne oppgaven beskriver utformingen av en botnet detektor og rapporterer resultatene fra testing av detektoren på reelle data fra en organisasjon i Norge. Det foreslåtte systemet er designet rundt et klassisk "misuse detection system". Det tar som input nettverksaktivitetslogg som NetFlow, DNS logg og HTTP logg og søker igjennom denne loggen med et stort signaturrett sett sammen av ulike signatursett som deles fritt på internett. Detektoren er basert på fire hovedkomponenter. 1) En algoritme for å kvantifisere risikoen representert ved en signatur, 2) En algoritme for hvitlisting av dårlige signaturer som vil skape falske positiver, 3) En søkemotor for å søke loggfiler med et stort signatursett, og 4) En algoritme for å identifisere kompromittert datamaskiner ved å aggregere alarm data.All komponentene sett under ett ser ut til å gi en betydelig forbedring i forhold til inntrengningsdeteksjon basert på vanlig signatursøk. En av de viktigste forbedringene er at systemene gjør det mye enklere å håndtere dårlige signaturer som skaper mange falske positiver, forbedring synes å være en kombinasjon av hvitlisting av noen av de dårlige signaturene og at fokuset flyttes fra arbeide med alarmer direkte til å jobbe med aggregert klient risikoer.Systemet er komplementære og synergistiske til noen av de nylig foreslåtte system i forskningslitteraturen som Exposure(Bilge, Kirda, Kruegel, & Balduzzi, 2011) og Notos (Antonakakis, Perdisci, Dagon, Lee, & Feamster, 2010)

Modern Graphics Processing Units (GPU) exhibit a high degree of parallelism and over the years have grown to adopt an increasing number of techniques to speed-up photorealistic rendering. One such technique is tessellation, i.e. the recursive subdivision of object elements into finer or coarser parts with the aim of achieving the appropriate amount of detail. The aim of this thesis is an adaptive tessellation algorithm for terrain rendering which can highlight important parts of the terrain while maintaining reasonable performance. This algorithm needs to be crack-free to avoid pixel faults between adjacent patches. A prototype was created in OSG, with GLSL shaders. The adaptive tessellation used three tessellation selection factors based on distance, a tessellation map and normals. An OSG program with five GLSL shaders (vertex, tessellation control, tessellation evaluation, geometry and fragment) was created for the tessellation of the terrain. The greatest advantage of tessellation is the reduced bandwidth between memory and GPU. Tessellation improves FPS, because it's faster to control vertices on the GPU than on the CPU.

Evolution in materio aim at revealing computational propertiesin materials. A bottom-up approach combining methods and theoriesfrom evolutionary algorithms, complex systems and unconventionalcomputation, is used to explore inherent computationalproperties of alternative computational materials. This master'sthesis build on earlier and ongoing research at NTNU focusingon interfacing materials. Mecobo is a prototype interface platform,including a physical interface, hardware for stimuli/measurementsand software, developed at NTNU.The Mecobo platform is intended to serve as a platform forexperimenting with unconventional materials' behavior in a dynamiccomplex system. This project describes the existing Mecoboplatform, modications to it and the new extension Cellular AutomataCentral Processing Unit (CA CPU). With the new extension,Mecobo is now capable of modelling dynamic complexsystems with a cellular automata and an unconventional material.The Cellular Automata Central Processing Unit extension is aparallel processor with 32 cores, but supports up to and including64 cores, for controlling each material conguration pin. Thecore is implemented as a stack machine optimized for instructionspace. It uses an 8-bit reduced instruction set computing (RISC)instruction set architecture (ISA).The modied Mecobo platform is tested and conrmed towork with several cellular automata rules and a simple modelconsisting of a theoretical material with constant response.

Porting a computationally demanding CUDA application to a GPU designed for mobile phones and tablets, which supports OpenCL, is the subject of this thesis.Significant effort is made to prepare the snow simulator of the HPC-LAB at IDI, NTNU, for porting to an OpenCL capable GPU for mobile phones, with a reasonably limited effort, when it arrives. The snow simulator is ported to OpenCL, documented, and improved by considering multiple sorting algorithms, as well as sorting the snow particles.A thorough study of GPUs for mobile devices and high performance computing, as well as their history is conducted to serve as a background for future porting of the simulator.The core code resulting from the OpenCL port is documented in detail to prepare for future projects on completing the port to a mobile device.The OpenCL port of the snow simulator is tested on a range of different OpenCL implementa- tions. The performance of GPUs designed for different use is compared, and memory manage- ment is identified as the biggest bottleneck for performance. This bottleneck is further investi- gated by studying the performance of the simulator when disabling certain copy operations.OpenCL supports more than just GPU devices. CPUs, Cell processors, and other acceleration cards are also supported. To investigate OpenCL on other devices, a part of the simulation is executed on a CPU, and compared with executing it on a GPU. The CPU version perform on par with the GPU version when using a laptop GPU.

Mobile gaming for smartphones has gained huge popularity in recent yearsand has become a big industry. In 2011 Google Play passed 10 billiondownloads. The game Angry Birds alone has been installed on devices morethan 50 million times.Increased performance of mobile devices means that new developmenttechniques can be used to make more advanced games. This projectpresents a new technique to develop multiplayer games using adistributed system.Multiplayer games often use a client-server architecture in order tocommunicate. Many games present a shared game universe to theplayers. Updates to the game state must be performedconsistently, otherwise discrepancies in the game state mayappear to the players.Mobile devices are unreliable servers for several reasons. First, theyuse a Wi-Fi connection that may not be reliable. Second, the deviceshave limited amounts of resources. The operating system may need toterminate the server, e.g., to recover memory for important servicessuch as receiving phone calls. Lastly, if the person with the serverchooses to stop playing, the game cannot proceed.A distributed system can prevent unreliable devices from causingfailures. When a device becomes unavailable, all its responsibilitiesshould be transfered to other devices in the system.This project presents the distributed system Same for sharingstate in mobile multiplayer games. Same is a peer-to-peer systemwith a master device that automatically recovers from failure. When amaster fails, the system may enter an inconsistent state. Recovery isperformed by electing a new master that restores consistency bycomparing the state of all the clients in the system. The system ispresented as a general model for sharing state and implemented for theAndroid operating system.Same is evaluated using benchmarks and an example application.Updates are performed consistently and distributed to all the connecteddevices. Same is shown to be a viable alternative to a standardclient-server architecture.

X-ray diffraction experiments are used extensively in the sciences to study the structure, chemicalcomposition and physical properties of materials. The output of such experiments are samples of thediffraction pattern, which essentially constitutes a 3D unstructured dataset. In this thesis, wedevelop a method for visualizing such datasets.Our visualization method is based on volume ray casting, but operates directly on the unstructuredsamples, rather than resampling them to form voxels. We estimate the intensity of the X-raydiffraction pattern at points along the rays by interpolation using nearby samples, taking advantageof an octree to facilitate efficient range search. The method is implemented on both the CPUand the GPU.To test our method, actual X-ray diffraction datasets is used, consisting of up to 120M samples.We are able to generate images of good quality. The rendering time varies dramatically, between 5 sand 200 s, depending upon dataset, and settings used. A simple performance model is developedand empirically tested to better understand this variation. Our implementation scales exceptionallywell to more CPU cores, with a speedup of 5.9 on a 6-core CPU. Furthermore, the GPU implementationachieves a speedup of around 4.6 compared to the CPU version.

NUTS er en 2U CubeSat, planlagt for oppskyting i 2014. NUTS er utviklet ved NTNU, og studenter ta del det å bygge og designe satellitten. Miniatyrsatellitter som følge Cubesat spesifikasjonen er ofte sammensatt av separate moduler. En databuss er vanlig metode for inter-modul kommunikasjon. Men å være en delt medium, trafikk på databusser følsomme for forstyrrelser fra sviktende moduler. NUTS bruker I2C som busstypen. Mangelen på sentraliserte voldgift positurer flere utfordringer med hensyn rettferdig bruk av buss tid, feil motstand og feilretting.Dette notatet beskriver prosessen med å utvikle en buss protokoll for NUTS. De viktigste målene for NUTS bussen protokollen er å gi noen grad av rettferdighet mellom modulene, diktere en nedre grense på data gjennomstrømming gitt visse forutsetninger, samt gi en nyttig abstraksjon å lette gjennomføringen av høyere nivå logikk. Flere støttefunksjoner for å gjøre dette mulig. Dette inkluderer å definere et system arkitektur og utvikle en grunnleggende sett med drivere som kreves for å kommunisere over en I2C buss.Hver modul deltar i utprøving av bussjåføren gjennomføringen, har USB-grensesnitt kontakter. Dette grensesnittet vil bli brukt til å kontrollere modulen eller moduler under test. Hvis noen ventetider skal ha garantert banen, må ventetid i avbruddet behandlingen være både forutsigbar og begrenset. Avbruddet behandling latency måles for å finne ut om dette kan være et problem.For å lette både testing og debugging, ble en USB-grensesnitt som brukes til å kontrollere de modulene under testing. Den eksisterende USB stabelen ble utvidet for å gi en egen kommunikasjonskanal, noe som gjør det mulig å gi terminal tjenester og dataoverføring funksjoner samtidig. Et design for NUTS bussen protokollen er også foreslått. For å hjelpe gjennomføringen av denne protokoll, har en I2C mester driver blitt utviklet. Et design for en $ I ^ 2C $ slavedriver har også blitt foreslåttVi finner ut at garantere rettferdig tilgang til en I2C buss krever spesielle hensyn i både sender og mottaker. Rettferdighet er et problem som må respekteres på system-nivå. Variasjoner i avbruddet behandling ventetid ved bruk FreeRTOS antyder at det kan være rom for forbedring.