Page 91 - Computer_Science_F5

Computer Science Chapter Performance and
Two

FOR ONLINE READING ONLY
optimisation of computer processor

Introduction

This chapter focuses on techniques that can be applied to improve the
performance of a processor and reduce access time. In this chapter, you will
learn the concept of pipelining, Instruction Level Parallelism (ILP), Thread
and data Level parallelism.

Knowledge regarding the performance and optimisation of computer processor
is fundamental in computer science, as it will equip you with the ability to
analyse processor performance bottlenecks, measure and quantify processors
performance, and parallelise programs/instructions effectively. The competencies
developed in this chapter will prepare you to design, build and use efficient,
cost-effective and high performing computer systems.

Think:
How a finely tuned engine boosts a car forward.

Pipelining

Concept of pipelining

Scenario 2.1: Kitchen as a simple processor

Imagine Mama Chapati, a popular stall known for its delicious flatbreads where
hungry students line up for their orders. However, during peak hours, the wait
times can be frustratingly long. Let us take a closer look at Mama Chapati’s
kitchen and see how it works, using the concept of pipelining to understand
how efficiency can be improved. Think of Mama Chapati’s kitchen as a simple
processor, with each step acting like a stage in the process. Assume each customer
orders only one chapati. Here is what happens in a non-pipelined approach:

82
for Advanced Secondary Schools

Computer Science Form 5.indd 82 23/07/2024 12:33