Computer Architecture

Introduction, Classes of computers, Defining Computer Architecture – Trends in Technology – Trends in Power and Energy in Integrated Circuits – Trends in Cost – Dependability – Measuring, Reporting and Summarizing Performance – Quantitative Principles of Computer Design.

Basic and Intermediate pipelining Concepts, The Major Hurdle of Pipelining – Pipeline Hazards, Pipelining Implementation, Implementation issues that makes Pipelining hard, The MIPS R4000 Pipeline, Extending the MIPS Pipeline to Handle Multicycle Operations.

Instruction-Level Parallelism: Concepts and Challenges – Basic Compiler Techniques for Exposing ILP – Reducing Branch Costs with Prediction – Overcoming Data Hazards with Dynamic Scheduling – Dynamic Scheduling – Hardware-Based Speculation – Exploiting ILP Using Multiple Issue and Static Scheduling, Advanced Techniques for Instruction Delivery and Speculation, Studies of the Limitations of ILP

Vector Architecture – SIMD Instruction Set Extensions for Multimedia – Graphics Processing Units – Centralized Shared-Memory Architectures – Performance of Shared-Memory Multiprocessors – Distributed Shared Memory, Models of Memory Consistency, Multicore Processors and Their Performance.

Review of Memory Hierarchy Design, Cache Performance, Basic Cache Optimizations, Virtual Memory, Protection: Virtual Memory and Virtual Machines, Advanced Optimizations of Cache Performance, Crosscutting Issues: The Design of Memory Hierarchies.