Moore's Law: A Complete Introduction

Moore’s law states that the transistor count on a microchip doubles every two years. This observation was first made in 1965 by Intel co-founder Gordon Moore. While it may sound hyperbolic at first, his estimation has held true for over 50 years. 

Let’s talk about how, and why, it matters.

(Related reading: Conway's Law & Amdahl's Law.)

How can transistor counts be doubling so quickly?

Moore’s observation was attributed to improvements in the chip manufacturing process where Intel was able to produce larger semiconductor components at scale with fewer defects. 

Specifically, the use of optical projection of lithography masks and fine-tuned image rendering on silicon wafers contributed to two-thirds of the improvement. 

Source: https://ourworldindata.org/moores-law

The last third is attributed to the optical doping process, which uses controlled light beams instead of applying chemical solutions to the wafer.

The accuracy of this prediction as demonstrated by a (logarithmically) linear increase in transistor count has sparked consumer interest and expectations. After all, the exponential growth trend of transistor density means that computing chips are getting faster and cheaper, and the consumer electronics market can experience a state of constant iteration and improvement.

A law of tech, or a self-fulfilling prophecy?

Major advances in manufacturing have made Moore’s Law ring true, but will it always be that way?

Moore’s law is not a fundamental law of physics: it does not describe a natural physical phenomenon. Noris it a scientific theory. Really, it’s an observation of how the semiconductor industry works. When taking a closer look, it seems that a combination of economic, technological and even social factors are driving this trend. 

The global semiconductor industry is worth over $0.5 Trillion dollars — a significant portion of this sum goes directly into research and development of new semiconductor technologies. The semiconductor industry has kept strong ties with academia, naturally attracting a concentrated effort among researchers around the world working toward improvements in semiconductor technologies. Similarly, manufacturing companies are large enterprises incentivized by governments pushing for supremacy in the competitive global landscape.

Another perspective of Moore's law states that it is a self-reinforcing goal that is driven by consumer expectations. Although this hypothesis is not validated or disproved, it is a realistic assumption that technology consumers appreciate two things:

• Regular and exponential improvements in technology performance 
• The constant reduction of cost for performance

Some argue that Moore’s law was claimed in an era of rapid technological innovations and advancements and that it may be reaching the end of its days as a “law”. Today, we have already reached a point where most chip manufacturing companies have struggled to improve transistor density from 7nm to 5nm transistor nodes in a highly competitive space.

Alongside these factors, there may be a philosophical aspect to Moore’s law. As the observation made decades ago has held true for so long, the semiconductor technology industry, academia and consumers perceive Moore’s law to be an attainable goal, despite increasingly glaring physical limitations. The collective vision appears to be a driving force of motivation that continues to propel the industry to achieve this goal — how long that can carry on remains to be seen.

Whatever the answer, Moore’s Law has rippling effects on businesses across industries. Understanding those effects might just help you stay ahead of the curve.

Let’s dig in.

Moore’s law for the enterprise

The observation gives remarkable insights into the progress of the semiconductors industry, leading to pervasive changes to the technological, economic and social domains across all industry verticals. Let’s discuss:

Accelerated technology adoption

Technology is evolving at a steady rate. 

Following Moore’s law, the semiconductor industry brings exponentially improving processing power to the market every two years. End users expect your business to take advantage of these improvements, which means that your organization should be prepared to adopt new digital transformation initiatives that streamline the digital experience for end-users. 

Improvements in transistor density plays an important role here: users expect shorter transition periods between technology updates, but the transition may well be characterized by increased risk. These risks include:

• System performance and dependability issues
• Steep learning curves for complex technologies
• Exposure to security issues such as zero-day exploits. 

Edge processing benefits and drawbacks

Over the last few decades, transistor density has increased to the point that a small form factor and low-power IoT device can pack sufficient computer power to enable edge computing. This capability allows business organizations to embed intelligence into the physical layer via sensor devices, and process data directly at the network edge. Using edge processing instead of transmitting sensitive information over large networks might result in cybersecurity risk or latency delays in data communication. 

From a business perspective, this presents an opportunity as well as a challenge.

On one hand, you can create more complex and data-driven IoT applications using advanced sensor technologies, but doing so requires a complete revamp of the technology stack originally designed for centralized data processing and storage while frond-end devices are only used as “dumb” terminals.

Thriving cloud model and scalability

One of the driving factors for rapid cloud adoption is that the service model enables users to trade high CapEx with affordable OpEx. 

Following Moore’s law, the semiconductor industry continues to supply faster computing capabilities for the same form factor and competitive pricing, cloud vendors can consequently offer highly scalable computing resources at an affordable OpEx. Although Moore’s law doesn’t explicitly define this economic impact, cloud investment trends are increasing exponentially and business organizations perceive a positive outlook for their cloud investments.

As long as Moore’s Law holds true, there’s a strong chance that cloud platforms will scale alongside emergent improvements in tech.

(Be sure to also keep an eye on emerging cloud cost trends.)

Assessing Moore’s Law in the future

Moore's Law has had a profound impact on the semiconductor industry, driving exponential advancements in technology and shaping various aspects of our lives. While it is not a fundamental law of physics, it has proven to be an accurate observation for over 50 years, prompting continuous research and development in semiconductor technologies. As technology reaches smaller transistor nodes, challenges arise, and the industry faces questions about the sustainability of the law's exponential growth. 

Whatever the outcome, remaining adaptive and understanding consumer expectations is critical in remaining competitive.