456-0322/01 – Advanced Architectures of Computers (PAP)

State of the art in microprocessor architecture, and consequence architectures of computers, obtaining of view to the practical activities of computers, and continuity with computer architectures. The frequentant yields an overview in present, and future processor architectures for computers used in the both general as well as special applications, Practical computations of the effectivity of processing are supported by circuit model simulations.

The university courses "Advanced Architectures of Computers" brings an overview of architectures of universal processors and techniques of instruction processing applied in last ten years with perspective to near future in the first part of lectures. Development of computer architectures streams namely to processors with large memories on the chip, which are used for saving processed program on one hand, and for saving larger amount of data to the short length from execution processing unit. As a part of new trends, vector processors are introduced, which we can see in the supercomputers. Development of their architectures will make changes of the next chips of computers. Graphic processors, which are discussed in the second part of lectures, represent most frequently used class of specialised processors in actual moment. Structure of specialised processors is derived from algorithms of solved problems, and it is effectively implemented as dedicated circuitry. Explanation follows the development of graphic accelerators to the second generation of graphic processors, and riches the advanced architectures of raster graphic. The third part of the course lectures is oriented to the starting class of the multimedia processors, the complexity and functional possibilities of which are limited by technologic bounds of implementation still; on the other hand they are implemented as multi-chip solution in this moment. Very high demands for the number of executed instructions of type calculation of convolution multiplication are repeated with high occurrence, which is necessary due to sampling frequency of multimedia signal. Substantial part of the third chapter is devoted to compression and error-control coding as well as cosine and wavelet transformation and to other methods applied in multimedia data processing.

Hennessy, J., Patterson, D.: Computer architecture Morgan Kaufmann Publishers (2003), ISBN 1-55860-596-7

Průběžná kontrola studia: Zpracování semestrálního projektu