Course Description

• Computer performance analysis - problems, techniques and areas of application.
• Performance evaluation of processors.
• Performance indicators of memory systems.
• Memory throughput. Memory utilization. Memory hierarchy. Effective address space.
• Performance evaluation of disks (linear, discrete and continuous models).
• Performance models of CD ROMs and magnetic tapes.
• Computer system modeling based on queuing networks.
• Poasson process. Birth-death process. Exponential and non-exponential models. Pollaczek-Khinchin formula.
• Closed networks. Cyclic model of multiprogramming. Central server model. Gordon-Newell's equations.
• Buzen's algorithm.
• Open networks, Jackson's theorem. Equivalent transformations (serial and parallel connection, cyclic connection).
• Interactive systems, deterministic and stochastic modeling.
• Operational analysis - open networks, closed networks, central-server networks, MVA algorithm.
• How to choose the most suitable technique for computer system performance analysis (analytical models, measurement techniques, simulation). Presentation and interpretation of results.