|
|
Decompilation of Specialized and Advanced Instruction Sets
Holub, Juraj ; Kolář, Dušan (referee) ; Křivka, Zbyněk (advisor)
V dnešnej dobe je proces analýzy nebezpečného softvéru dôležitou súčasťou informačných technológií. Jedna z kľúčových techník je spätný preklad škodlivých binárnych programov. Spätný preklad je komplexný proces, ktorý rieši niekoľko projektov. Projekt RetDec sa zameriava na flexibilný návrh a riešenie spätného prekladača s možnosťou znovupoužiteľnosti. Cieľom tejto práce je zlepšenie spätného prekladu pokročilých inštrukčných sád pre architektúru x86. Bola navrhnutá nová optimalizácia pre FPU registrový zásobník. Bola rozšírená podpora prekladu inštrukčných sád jednotiek FPU a SSE. Nové rozšírenia boli implementované a otestované z hľadiska efektivity a kvality spätného prekladu.
|
|
|
Native Code Web Browser Extensions
Vítek, Vojtěch ; Očenášek, Pavel (referee) ; Burget, Radek (advisor)
Nowadays, web applications and browsers are undergoing rapid development - we can say that the progress of Internet technologies is unstoppable phenomenon of the last decade. The need for the best-possible CPU performance in web applications to achieve a smooth user experience is enormous - besides the continuous improvements of existing technologies, we can see several new technologies arising every year. This thesis deals with development of native code web browser extensions whose primary purpose is to use maximum CPU performance as well as efforts to improve the user experience when viewing web pages.
|
|
| |
|
|
Intel Integrated Performance Primitives and their use in application development
Machač, Jiří ; Přinosil, Jiří (referee) ; Malý, Jan (advisor)
The aim of the presented work is to demonstrate and evaluate the contribution of computing system SIMD especially units MMX, SSE, SSE2, SSE3, SSSE3 and SSE4 from Intel company, by creation of demostrating applications with using Intel Integrated Performance Primitives library. At first, possibilities of SIMD programming using intrinsic function, vektorization and libraries Intel Integrated Performance Primitives are presented, as next are descibed options of evaluation of particular algorithms. Finally procedure of programing by using Intel Integrated Performance Primitives library are ilustrated.
|
|
| |
|
|
The Efficient Implementation of the Genetic Algorithm Using Multicore Processors
Kouřil, Miroslav ; Žaloudek, Luděk (referee) ; Jaroš, Jiří (advisor)
This diploma thesis deals with acceleration of advanced genetic algorithm. For implementation, discrete and continuos versions of UMDA genetic algorithm were chosen. The main part of the acceleration is the utilization of SSE instruction set. Using this set, the functions for calculating fitness and new population sampling were accelerated in particular. Then the pseudorandom number generator that also uses SSE instruction set was implemented. The discrete algorithm reached the speed of up to 4,6 after this implementation. Finally, the algorithms were modified so that the system OpenMP could be used, which enables the running of blocks of code in more threads. The continuous version of algorithm is not convenient for parallelization, because computational complexity of that algorithm is low. In comparison, the discrete versions of algorithm are really appropriate for parallelization. Both the implemented versions reached the total acceleration of up to 4,9 and 7,2.
|
|
|
Information System Assesment
Menšík, Marek ; Ing.Kamil Macuch (referee) ; Koch, Miloš (advisor)
This diploma thesis deals with information systems evaluation from user and producer aspects. The theoretical part presents current trends and actual ways of using, including practical application used in real firms. The practical part is focused on proposals and changes for concrete producer of information system.
|
|
|
Selected Fractal Algorithms and Their Use
Michálek, Bedřich ; Španěl, Michal (referee) ; Herout, Adam (advisor)
To read up on the problems of fractal algorithms, to program a simple demo application of this algorithms and propose alternatives of a further study was the goal of my project. Since the subject of fractals and fractal algorithms is very extensive, I concentrated on three specific types of them - the Mandelbrot set and the Julia sets related to it, iterated function systems with application of different algorithms for generating, and the stochastic fractals and their utilization for grass, shrub and plasma modelling. I tried out for a certain rate of their optimalization, acceleration and the comparison of time and memory requirements in my study and application although it is not ever possible due to the characteristic of the given fractal type. I also theoretically present some utilization options of the SSE technology for the acceleration of fractal algorithm calculations. The demo application is proposed in the very easy way because there was not the goal to programme a new carefully worked-out software, but rather to try to implement the read up algorithms. For this reason I focused on the clear implementation of these algorithms to be their utilization easily possible. The application is written in C/C++ language with SDL library used to handle the graphics output.
|
|
|
Assisted Code Vectorization and Parallelization Using the OpenMP 4.0 Standard
Slouka, Lukáš ; Nikl, Vojtěch (referee) ; Jaroš, Jiří (advisor)
The subject of the bachelor's thesis is code optimization using the OpenMP 4.0 standard which provides tools for assisted parallelization and vectorization. In addition to the descrip tion of the OpenMP 4.0 standard, the thesis as well contains an insight into architectures of modern computers, specifically the system of cache memories and SSE/AVX modules that play a major role in the optimization field. The thesis demonstrates advantages of optimized code compared to unoptimized version on a set of benchmarks which are aimed at various aspects of optimization.
|
|
|
Application for Alignment of DNA Parts
Kašpárek, Tomáš ; Žák, Jakub (referee) ; Rozman, Jaroslav (advisor)
This thesis deals with DNA sequences alignment methods with focus on execution speed of given task and its optimality. The outcome of this thesis results in multiple programs, which presents DNA alignment algorithms and their versions programmed with OpenCL libraries, which are optimized on calculation rate. Text of this thesis apprises reader with DNA alignment problematic and its importance in biology. Further, algorithms for DNA alignment and options of their speed-up using libraries like CUDA or OpenCL are presented.
|
guest ::
