The Multiple Lives of Moore’s Law

Mack, C. (2015). Moore’s Law Has Many Lives. IEEE Spectrum: News in Technology, Engineering, and Science.

http://spectrum.ieee.org/semiconductors/processors/the-multiple-lives-of-moores-law/retrieved 12 May 2017.

Bibliographic

References

Laws, D. (2014). Who Created the IC? The Museum of Computer History. Computerhistory.org. http://www.computerhistory.org/atchm/who-invented-the-ic/ (accessed 12 May 2017).

J. Y. C. Sun (2014). Semiconductor innovation will continue into the next decade. In the 2014 IEEE Asian Solid-State Circuits Conference (A-SSCC) (pp. 117-120). doi: 10.1109/ASSCC.2014.7008874

Objectives

Explains why Moore’s great prediction has stood the test of time for almost a half-century.

Discusses the advancement of integrated circuits.

reveals information about Moore’s prediction that are frequently neglected

Shows how Moore’s law is about to change.

Shows Moore’s prediction will endure for many years to come

Summary

About a half a century ago, Gordon E. Moore, in his article Electronics predicted a future with home computers, smartphones and automatic control systems for automobiles. He argued that the development would be driven by the increasing number of circuit components that could be integrated into a single chip. Mack (2015) agrees that today’s computers, sensors, personal electronics and other necessities draw a lot from the prediction. Moore’s law is a complex concept whose meaning has changed over the years.

In the 1960s, Silicon Valley was among the few companies that were making a significant contribution in the field of transistors. At that time, circuits were mostly constructed from individual capacitors, resistors, transistors and diodes that were manually wired on a circuit board. After a thorough research, Silicon Valley discovered that planar transistors could make an integrated circuit as a single chip. Although the idea was faced with skepticism then, the future electronics seems will rely on the success of integrated electronics.

Integrated circuits were initially compact but more expensive than the hand-wired. In addition to that, the transistors were unreliable and as a result, only a few companies especially those working with government agencies like NASA and the US military were making the integrated circuits. The author argues that the limitation was addressed by making the chip with multiple transistors; the approach produces only a fraction of working chips similar to what would be obtained from the same number of standalone transistors.

Further, the paper reveals that Moore’s prediction was not only about transistors but was also concerned the number of electronic devices such as diodes, capacitors, and resistors. Moreover, Moore understood that the numbers of components that could be packaged on a chip and that could make an economic sense were different. Ideally, economies of scale dictate that adding more components reduces the cost per part. It is in this light that the development in chip-fabrication technology has been seen to divert its focus to producing more components at lower costs per chip.

In recent years, shrinking transistors while maintaining their speed and power consumption has become increasingly difficult. Accordingly, Moore’s law has shifted its focus to cost rather than performance. The economics appeal has remained relevant because making the transistors smaller today can be achieved at the same cost of manufacturing each unit of silicon. Keeping the cost of silicon constant for decades has been made possible by the technological improvements that date back in the 1970s. However, the article predicts that the trend is on the verge of evolving given that lithography is becoming more expensive. Even though the new rendition of the law may not make meaningful economic sense, the author points out that it will be better than the previous.

Results

Moore’s Law is not evidence of technology’s inevitable march but a proof of human ingenuity, hard work and the incentives of a free market

The meaning of Moore’s law has repeatedly changed over the past a half century

Moore’s prediction is concerned with the number of electric components and economics of integration

Moore’s prediction is currently changing because of the benefits of miniaturizations are deteriorating

Better forms of integration will define progress in the new rendition of Moore’s law

Critique

First, the title of the article is relevant given the different phases that Moore’s law has undergone over the years. The law has endured 50 years of transformation characterized by two stages of development. According to Mack (2015), the first phase of law included scaling up of the number of components being integrated. This step was followed by decreasing the size and cost of transistors. These changes and the fact that the law is still changing shows that the prediction has lived multiple lives.

Although the purpose of the article is not explicitly stated, it can be seen from the introduction paragraph that paper investigates the steady doubling predicted by Moore. So far, the doubling has been witnessed in the increasing number of components on integrated circuits. Laws (2014) agrees that most semiconductor manufacturers have opted for the planar design that supports today’s billion-transistor chips. Indeed more exciting developments in this field are expected yet to come.

Mack’s (2015) argument that innovations in semiconductors will continue is sound. The fact that Moore’s Law is morphing is clear indication of the predicted trends. In addition to that, another recent research shows that the technology will continue to create an indispensable infrastructure for many industries. Semiconductors, especially in silicon-based electronics, will continue to improve virtually every aspect of human life.

Questions

What is the implication of Moore’s law today?

In my opinion, the law has far-reaching consequences in the development of cloud computing and social media technologies. These inventions require increased computer capacity and are responsible for the high demand for the integration of more components in a single chip.

Is the future of integrated electronics the future of electronics?

The integration of different electronic components on a single chip has made significant improvements in electronics regarding size, performance, and power consumption. For instance, it has caused the migration from microelectronics to Nano-electronics and hence the emergence of nanotechnology. Indeed, the integration of components holds the future for electronics.

How can the leakage of energy caused by miniaturization be prevented?

I think the problem can be addressed using high-k dielectrics instead of SiO2. For instance, a compound like HfO2 can be used because its dielectric constant is six times larger than that of SiO2.

Which other problems are caused by miniaturization besides the dielectric leakage?

Reducing the channel length and width can cause the small geometry effects including surface scattering, hot electrons, and velocity saturation, to name but a few. Miniaturization also increases the parasitic series of resistance of the transistor, which in turn reduces the efficiency of the chip. Lastly, the effect of randomness of distribution of the chip properties has become severe in small devices because there exist about 100 dopant atoms in a 50nm transistor.