Course: Computers Architecture III

« Back
Course title Computers Architecture III
Course code UAI/606
Organizational form of instruction Lecture
Level of course Master
Year of study not specified
Frequency of the course In each academic year, in the winter semester.
Semester Winter
Number of ECTS credits 4
Language of instruction Czech
Status of course Compulsory, Compulsory-optional
Form of instruction Face-to-face
Work placements This is not an internship
Recommended optional programme components None
Lecturer(s)
  • Skrbek Miroslav, Ing. Ph.D.
  • Šebesta Václav, doc. Ing. DrSc.
Course content
Lecture topics: 1. Von Neumann model. 2. Flynn taxonomy of parallel systems, types of parallelism. 3. Amdahl law, efficiency of parallel computation, performance measurement, benchmarks. 4. tight and loosely coupled parallel architectures, topology of connections, processor-to-memory connections, switches. 5. Instruction set architectures (ISA). 6. Scalar processors, CISC and RISC architectures. 7. Principles of pipelined data and instruction processing, hazards, code optimization. 8. Out-of-order and speculative instruction processing. 9. Memory subsystem, virtual memory, memory consistency. 10. VLIW and SIMD architectures. 11. Symmetric multiprocessing, multi-core symmetric and hybrid architectures. 12. Graphics accelerators. 13. Neural chips, neural accelerators.

Learning activities and teaching methods
Monologic (reading, lecture, briefing)
  • Semestral paper - 40 hours per semester
  • Class attendance - 28 hours per semester
  • Preparation for exam - 40 hours per semester
Learning outcomes
The aim of the course is to teach students a wide range of computer architectures starting from basic von Neumann model, scalar and superscalar processors through a tight and loosely coupled parallel architectures to massively parallel and application specific architectures. In detail, it focuses on RISC architectures, their design and principles that are used in modern processors. The course shows the instruction-level, processor-level and system-level parallelism including software problems parallelization, synchronization and code optimization. It includes problems of computer performance, efficiency of parallel computation, performance measurement and benchmarking.
In this course, students obrain deeper knowledge in the field of the computer architectures, especially in paralelism at the instruction level paralellism and the processor level.
Prerequisites
Knowledge of computer architectures at bachelor degree level (UAI698 a UAI682). Programming in Java.

Assessment methods and criteria
Combined exam, Seminar work

Each student may take 100 points (70 points examination, 30 points project,report, presentation). For passing examination, the total number of points (examination and tutorial) must be greater or equal to 50, the examination test must be evaluated to one half points or more and the project must be evaluated to 10 points as minimum. If any of these conditions is not satisfied, the student fails.
Recommended literature
  • Dvořák,V.,Drábek,V. Architektura procesorů. Brno, 1999. ISBN 80-214-1458-8.
  • Hennessy,J. L., Patterson, D. A. Computer Architecture: A Quantitative Approach. San Mateo, CA, Morgan Kaufman, 2002. ISBN 1-55-860596-7.


Study plans that include the course
Faculty Study plan (Version) Category of Branch/Specialization Recommended year of study Recommended semester
Faculty: Faculty of Science Study plan (Version): Secondary Schools Teacher Training in Informatics (1) Category: Pedagogy, teacher training and social care - Recommended year of study:-, Recommended semester: Winter
Faculty: Faculty of Science Study plan (Version): Applied Informatics (1) Category: Informatics courses - Recommended year of study:-, Recommended semester: Winter