Modern computer applications often need to get the maximum performance out of the computer. Whereas advances in hardware technology mean that the speed of the CPU increases by about 50% a year, corresponding developments in operating systems and other system software lag behind.
This course is concerned with qualitative and quantitative techniques used in the construction and evaluation of operating systems for computer systems where performance is important, such as real-time applications, communication systems and multimedia systems, or in systems which have to perform huge calculations or use very large amounts of data. In particular, we consider quantitative and algorithmic aspects of real-time scheduling, virtual memory and memory management, task management and synchronisation in uniprocessor and multiprocessor systems, inter-process communication, drivers and I/O systems, disc and file systems and networking.
The exact content of the course varies from year to year. In 2004 the emphasis will be on cluster architectures and Grid computing.
Course description
Teaching material
Lecture plan
Exercises
End of term projects instead of written exam
telephone: 45 25 37 49 e-mail: robin@imm.dtu.dk