CS704 Handouts 2023

Advance Computer Architecture II

Course - Code: CS704

Reference by Computer Architecture, A Quantitative Approach by Jhon L. Hennessy and David A. Patterson

Chapter · Subject · Lectures Page

Fundamentals of Computer Design 01 - 03

(Appendix B) Instruction Set Principles 04-06

(Appendix A) Basic and Intermediate Concepts 07-11

2 - 3

Instruction-Level Parallelism and Its Exploitation 12-24

25-33

Multiprocessors and Thread-Level Parallelism 34-37

Storage Systems 38-45

Presented by:

Dr. M. Ashraf Chughtai

Ph.D, University of Manchester, UK

Reference Book 1:

Computer Architecture, A Quantitative Approach by Jhon L. Hennessy and David A. Patterson

Download Link:

Reference Book 2:

Computer Organization and Architecture: Designing for Performance by William Stalling

Download Link: · A World Class Education at Your Door Step · Lecture · Subject · Page · Quantitative Principles

Detailed discussion on the computer Performance Design for Performance

Encoding instructions and MIPS Instruction format ISA Performance Analysis, Fallacies and Pitfalls

Computer Hardware Design

Single Cycle Data path and Control Design Multi Cycle Datapath and Control Design Multi Cycle and Pipeline - Datapath and Control Design Pipeline Datapath and Control Design Pipeline and Instruction Level Parallelism

Introduction to multi cycle pipelined Datapath Dynamic Scheduling - Scoreboard Approach Dynamic Scheduling – Tomasulo‘s Approach Dynamic Branch Prediction Dynamic Branch Prediction …. Cont‘d High-performance Instructions delivery - Multiple Issue Hardware-based speculations and exceptions Limitations of ILP and Conclusion Static Scheduling Static Scheduling – Multiple Issue Processor Software pipelining and Trace Scheduling Hardware Support at Compile Time Concluding Instruction Level Parallelism

Storage Technologies Trends and Caching Concept of Caching and Principle of Locality Cache Design Techniques Cache Design and policies Cache Performance Enhancement by: Reducing Cache Miss Penalty Cache Performance Enhancement: Reducing Miss Rate Cache Performance Enhancement by: Miss Penalty/Rate Parallelism and Hit time Main and Virtual Memories Main and Virtual Memories …. Cont‘d

Shared Memory Architectures Cache Coherence Problem Cache Coherence Problem … Cont‘d Performance and Synchronization

Input Output Systems

Storage and I/O Systems Bus Structures Connecting I/O Devices RAID and I/O System Design

Networks and Clusters

Networks: Interconnection and Topology Networks Topology and Internetworking ... Cont‘d Internetworks and Clusters Putting It All Together: Case Studies Putting It All Together: Review: Lecture 1 - 43

Lecture 1

Welcome to the course of Advanced Computer Architecture. This course of Advanced Computer Architecture has been developed with the assumption that the students have basic knowledge of: digital logic design, computer organization and design, programming model of microprocessor; memory and input/output interfacing with the microprocessor and fundamentals of Computer Architecture My name is Muhammad Ashraf Chughtai, After completing my Bachelor‘s and Masters degrees from the University of Engineering and Technology, Lahore in 1974 and 1978, respectively I completed my Ph.D. from the University of Manchester (UMIST) UK in 1986. After spending 32 years at UET Lahore, as lecturer through professor, presently I am associated with COMSATS.

Computer System

Computer system is a collection of Central Processing Unit, memory system and peripheral devices, all interconnected by groups of conductors called buses

Computer Architecture Verses Organization

Architecture refers to those attributes of a computer visible to a programmer or compiler writer such as: instruction set, addressing techniques, I/O mechanisms etc. The members of each family (say Intel x86 family and IBM family) share the same basic architecture and have code compatibility, Organization refers to how the features of a computer are implemented; i.e., control signal generation as FSM or microprogramming, memory technology-SRAM, DRAM etc, hardware or software based realization of operation-multiplication by hardware or algorithmically. Moreover, the organization of same architecture may differ between different versions; i.e., different versions of Intel x86 family may have different organizations

Academic History:

In 1944, John von Neumann introduced the concept of stored-program computer, referred to as Electronic Discrete Variable Automatic Computer – EDVAC.