An ideal computer is quick with processing (around 4 GHz which requires access times of a few nanoseconds) and has a large memory (some terabytes). Unfortunately, large memories are usually slow (some milliseconds access time) and fast memories (a few nanoseconds) are expensive. It needs not only efficient storage setup concept but also efficient processors to master this balancing act. In module “processor architecture” students get more familiar with how processors conceptionally look like and how an efficient memory is structured.

For this purpose, the module "processor architecture" merges knowledge from other lectures and forms a complete image out of this. The lecture discusses an exemplary processor step by step first. Afterwards, the memories are discussed. Fast memories must be kept small for saving costs. For providing all data at any time to the processor despite of that, different concepts up to hybrid solutions from software and hardware must be applied.

The module lasts one term. Lectures, tutorials and practical trainings are attended here. The students learn which criteria characterize an efficient processor and how a processor’s performance can be measured. Furthermore, they develop a feeling for the consequences of design decision. For developing a modern processor, an understanding for machine codes and assembly languages is needed. The students learn to advantageously utilize the machine code characteristics. During the practical training, they acquire the related practical assembly programming skills.

The module "Processor architecture" was a highlight of my previous studies. It combines knowledge from nearly all modules for elaborating a complex topic. The interaction of hardware and software is impressive and enables solutions that otherwise could only be implemented with difficulties.