Moore's Law Related to Google

The end of MooreÐŽ¦s Law?

Table of Contents

1. Introduction 2

2. MooreÐŽ¦s Law 3

3. Stakeholders 5

3.1 Companies 5

3.2 Governments 5

3.3 Academics 6

3.4 End-users 6

4. Driving Forces 7

4.1 Self Fulfilling Prophecy 7

4.2 Technical Possibilities 7

4.3 Economic Aspects 8

5. Scenarios 9

6. The consequences of the failing law for the main stakeholders 10

6.1 The governmental (macro) domain 10

6.2 The corporate (meso) domain 10

6.3 The end-user (micro) domain 12

6.4 The academic domain 12

7. Contingency Strategies to reduce Risk Exposure 13

7.1 The governmental (macro) domain 13

7.2 The corporate (meso) domain 13

7.3 The end-user (micro) domain 13

7.4 The academic domain 13

8. Conclusion 14

9. References 15

1. Introduction

We live in a period where technology is developing fast and is causing competition between many companies and organizations. Personal computers are getting better, faster and are more reliable, which increases the development of new technologies in different areas. New operating systems and complex software are developed using these ЎҐsuperÐŽ¦ computers, increasing the need for even better PCs that can operate these demanding applications. The ability to continue developing better central processing units (CPUs) and better PCs is an issue that affects many people and companies; this brings up the issue of the survival of MooreÐŽ¦s law.

MooreÐŽ¦s law discusses the fact that the number of transistors on an integrated circuit doubles every 24 months, and that this results in better computers that give the possibility for the technology to keep growing in a fast moving track. This law was created in 1970 and is attributed to Gordon E. Moore. Many opinions which question its validity have popped up ever since.

This paper discusses different aspects of MooreÐŽ¦s law, tries to capture a forecast of the future of this law, and discusses the consequences if MooreÐŽ¦s law would come to an end. We first describe MooreÐŽ¦s law and the changes in the law that occurred since its creation. We will discuss the self-fulfilling prophecy viewpoint, and the technical, physical, and economical obstacles that raised the question of validation of this law.

Further discussion will be about the alternatives and different scenarios if MooreÐŽ¦s law comes to an end, what other technologies can be used to improve computer performance, and about the continuation of a steady increase in technology.

We dedicated a full section for the general consequences if MooreÐŽ¦s law will not be valid anymore. We will explain the effects of this on the economy and industry, and the reasons for those effects. We will also focus on the consequences for the search engine market and other related markets in which Google operates. The question about the gains and losses if MooreÐŽ¦s law will come to an end and the effect on the stakeholders will be answered in this paper. This will give us a broader view of the different stakeholders in the computer market and the common worries that the industry shares.

2. MooreÐŽ¦s Law

In 1965, the co-founder of Intel, Gordon E. Moore said that ÐŽ§the number of transistors on an integrated circuit for minimum component cost doubles every 24 monthsЎЁ. This idea was a useful predictor for the rapid development in computer technology since the number of transistors on a chip indicates the computer processing power. Because the idea seemed to be more or less valid, it was named in 1970 ÐŽ§MooreÐŽ¦s LawЎЁ. Figure 1 shows the relation between the exponential growth of the number of transistors and the steady increase in performance of the Intel processors since 1970.

Figure 1: MooreÐŽ¦s Law

There have been several variations of MooreÐŽ¦s Law since its creation, for example: the number of transistors will grow exponentially every 12 or 18 months. While these variations affect the development pace, they do not change the idea and the conclusion you can withdraw from this law and the impact it created.

When MooreÐŽ¦s law became familiar, opinions came along and criticized this law. How long this law can exist, and how long it takes until it will not be possible to add more transistors onto the integrated circle were the main critics. The computation complexity theory stated clearly that this kind of exponential behaviour is not feasible in practice for infinity. This means that the law was feasible in the past and still feasible in the present, but one day MooreÐŽ¦s law will come to an end.

Just to demonstrate this powerful law and the concept it entails, assume we take a problem that needs 2n operations to be solved, where n is the size of the input (this is an exponential problem exactly as MooreÐŽ¦s law stated). Taking a small value for n, for example n=100 will result in 2100 operations that need to be performed to solve our problem. These days, computers can perform 1010 operation per second which means that it will take 4*1012 years to solve our problem, which is longer than the age of the universe. Therefore, practically, this law is not feasible and will be terminated in the future.1

PCs and operating systems are interrelated and related to MooreÐŽ¦s law. These days, a desktop PC is not just a tool for word processing and simple calculation. It is a powerful machine that performs millions of operations per second, can be used as a network server, and