Computer architecture

About this programme

No fees are charged for EU and EEA citizens, Swedish residence permit holders and exchange students. The course covers architectural techniques necessary to achieve high performance for user programs. It also covers simulation-based analysis methods for quantitative assessment of the impact of a particular architectural technique on performance. 1. The first part covers aspects affecting the development of computer technology including Moore's Law, performance measures (execution time vs. throughput), benchmarking, and fundamentals of computer performance such as Amdahl's Law and memory access locality. 2. The second part covers various techniques for exploiting instruction level parallelism (ILP) by defining key concepts of what ILP is and what limits it. The techniques covered fall into two categories: dynamic and static techniques. The main dynamic techniques covered are Tomasulo's algorithm, branch prediction and speculation. The most important static techniques are loop unrolling, software pipelining, trace scheduling and predicated instructions. 3. The third part covers principles of modern memory hierarchies such as inclusive and non-inclusive memory hierarchies, lockup free caching techniques, prefetching, classification of cache misses. 4. The fourth part deals with multi-core microprocessors and techniques such as multi-threading and cache coherence. Requirements to participate in the course are that two years have been successfully completed in computer science or equivalent. The course DIT122 (Computer System Engineering) or equivalent is a prerequisite.