Categories
Latest
Popular

The End of Moore’s Law Is Coming. Here’s What You Need to Know

By Wgsimon (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

By Wgsimon (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

If you’re a fan of technology, you’re likely familiar with Moore’s Law. You’re also probably aware that it is set to become inapplicable soon—sooner than expected. After several decades of advancements in the manufacture of transistors, the limitations have finally made their onset. As EE Times reports, 28nm will be the last node of Moore’s Law.

For those who still haven’t heard the news, the following info bits should put you on track.

Overview of Moore’s Law

Moore’s Law is an observation in computing hardware named after one of the founders of Intel, Gordon E. Moore. It was first published in a paper in 1965. It is mainly about the tendency in computing hardware for the number of transistors used on integrated circuits to increase by two times in about every two years. So far, this observation has been proven to hold true.

Additionally, as published in Moore’s paper entitled “The Future of Integrated Electronics,” there is a cost aspect in Moore’s Law. Below is an excerpt of the text:

“The complexity for minimum component costs has increased at a rate of roughly a factor of two per year. Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years.”

Moore’s Law has been used in the semiconductor industry to provide guidance in planning long-term hardware manufacturing goals and in conducting research and development efforts. Moreover, developments in the production of many electronic devices over the years can be associated with Moore’s Law.

Guiding light at English Wikipedia [CC-BY-SA-3.0-2.5-2.0-1.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

Guiding light at English Wikipedia [CC-BY-SA-3.0-2.5-2.0-1.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

Not Exactly a Law

Unlike the Law of Gravity or the Law of Energy Conservation, Moore’s Law is not really some scientific law. It is merely a conjecture or an observation that somehow held true for more than five decades. It has somehow become a defining phrase for electronics manufacture in the later part of the 20th century and early 21st century.

Why Is Moore’s Law Ending?

In 2005, it was predicted that Moore’s Law will cease to be applicable by 2015 or 2020. This prediction, however, was revised in 2010 as the International Technology Roadmap for Semiconductors anticipated slow growth by the end of 2013.

The reason for Moore’s Law becoming inapplicable at 28nm is, for now, not the limitation on how much smaller transistors can be made and how many of them can be packed in the same die size. It is still possible to make them smaller and more compact. The problem is in the cost.

While it is still possible for current technology to further miniaturize, the cost of doing so will already be higher and will go against what Moore’s Law has established.

As mentioned above, Moore’s Law is not just about the effort of making transistors smaller and more compact. There’s also a cost component. Several companies have prepared graphs and presentations that support the idea that Moore’s Law has already ceased to be applicable.

  • Joel Hartmann, Executive Vice-President of Manufacturing and Process R&D , Embedded Processing Solutions at ST, made a presentation on the discontinuation of Moore’s Law on account of cost stagnation or increase.

  • GlobalFoundries also have data supporting the idea that Moore’s Law is about or has already come to its end. They have an FDSOI Costs/Positioning graph that highlights how the lowest-cost transistor is limited to the polySION 28nm node. Any attempts to scale above it will already be very expensive.

  • Of note, wafer cost is also increasing, as shown by a chart from Semicon Japan. Beyond 28nm, cost already becomes a major hurdle for manufacturers.

By Hard_drive_capacity_over_time.png: Hankwang derivative work: Rentar (Hard_drive_capacity_over_time.png) [CC-BY-SA-3.0-2.5-2.0-1.0 (http://creativecommons.org/licenses/by-sa/3.0), GFDL (http://www.gnu.org/copyleft/fdl.html) or Public domain], via Wikimedia Commons

By Hard_drive_capacity_over_time.png: Hankwang derivative work: Rentar (Hard_drive_capacity_over_time.png) [CC-BY-SA-3.0-2.5-2.0-1.0 (http://creativecommons.org/licenses/by-sa/3.0), GFDL (http://www.gnu.org/copyleft/fdl.html) or Public domain], via Wikimedia Commons

Atomic Size: The Fundamental Barrier

Even if compromises are still available to lower costs, Moore’s Law is still expected to become inapplicable, albeit not in the next few years or decades. Gordon Moore himself expected the nearing end of Moore’s Law in an interview in April 2005. He said that “it can’t continue forever” and that transistors are bound to eventually reach the limits of miniaturization. Of course, a transistor cannot be smaller than an atom or even a molecule. The transistors being used at present are already approaching the size of atoms.

Of course, the end of Moore’s Law does not really create an impact on technology. Tech advancement will continue whether or not Moore’s Law remains applicable. It is simply an acknowledgment of how the times have changed, how we are nearing the limits of dimensional scaling and cost effective production in the prevailing market of electronics component supply.