The end of Moore’s Law – A History and what comes next

Moores Law Transistors Per Area 2015There’s a change happening in the semiconductor industry, with Moore’s law no longer being the guiding ideology. Could this mean the end of Moore’s Law? Let’s find out together.

Moore’s law started in the 1960s by Gordon Moore and kick-started the information-technology revolution that has grown ever since. Semiconductors make digital devices work, without them computers, smartphones and devices in the Internet of Things wouldn’t be possible.

Moore’s law states that microprocessor chips will have twice the number of transistors every two years. This couldn’t be infinite, but it meant that microprocessors improved drastically in a short space of time. For a period in the 1990s and 2000s, it felt as if computers were improving quicker than it was possible to upgrade them. That’s changed now, for example with smartphones. Smartphones were a revelation when released and improvements on functionality, processor speed and capability happened fast. Now, the high-end model smartphones don’t demand to be bought as they used to be and don ‘t have the massive upgrades that they previously had.

Semiconductors are part of what brought the internet into billions of hands, allowed information to be sent over long distances instantaneously and connected people across the world. Chip makers, such as Intel, Taiwan Semiconductor Manufacturing Company, and Samsung have been following Moore’s law to its end point. The smaller the semiconductors, the more semiconductors in the processor, the more heat that’s generated by the processor. The heat created means that microprocessors are unable to do their jobs properly.

Microprocessor currently has circuits that are 14 nanometres (To put that in perspective, most viruses are larger than these circuits). Getting smaller, down to the 2-3 nanometer limit, means quantum uncertainties come into play and make the device unreliable. Getting into that size is an unknown with what will happen with silicon technology. There’s been a lot of research into other materials beyond silicon, but nothing has been found yet.

Bill Bottoms, president of Third Millennium Test Solutions, “It used to be that whenever we would scale to smaller feature size, good things happened automatically. The chips would go faster and consume less power.” This stopped happening in the early 2000s, and as stated above the chips became too hot when the circuits became less than 90 nanometres.

The tech industry claims that the end of Moore’s law is not far away, but other industries show that there will be improvements to be made. With airplanes, the Boeing 707 from the 1950s travels at the same speed as the Boeing 787, but the improvements in other areas are pronounced. Applications could be built to use microprocessors better like the improvements made in airplanes. Add-ons like sensors and power management circuits can be developed to get better speeds and functionality out of silicon devices.

A roadmap for semiconductor improvement was overseen by the US semiconductor industry was launched in 1981 and became an international collaboration with the International Technology Roadmap for Semiconductors. Industrial associations in Europe, Japan, Taiwan and South Korea were involved. This year, a new roadmap is to be released, titled the International Roadmap for Devices and Systems.

According to Kenneth Flamm of the University of Texas who studies the economics of the computer industry, “The road map was an incredibly interesting experiment. So far as I know, there is no example of anything like this in any other industry, where every manufacturer and supplier gets together and figures out what they are going to do.”

The future of computing is possibly in a new paradigm of quantum or neuromorphic computing. Quantum computing promises exponential speed-ups when it comes to calculations and processing. Neuromorphic computing aims to build processors that process information close to how the brain processes neurons. Quantum computing is possibly going to be for niche applications rather than everyday applications. The sciences, manufacturing, and other industries will benefit from quantum specific computing. Neuromorphic computing seems like something that will help to bring in an artificial intelligence revolution. Neither quantum or neuromorphic computing has made it out of the testing labs as yet.

Heat is something that the semiconductor industry is working on. If there is a material that exists or can be invented that doesn’t heat up like silicon, then circuits could get smaller in size. Microprocessors are stacks and stacking elements that heat up on top of other elements that heat up ends up causing a lot more heat in the device.

Economics is another issue that Bottoms mentions when it comes to Moore’s law. The more that semiconductors shrink, the more they cost. At what point does the cost outweigh the benefits of a smaller device?

“The old market was that you would make a few different things, but sell a whole lot of them,” says Daniel Reed, a computer scientist and vice-president of research at the University of Iowa. “The new market is that you have to make a lot of things, but sell a few hundred thousand apiece — so it had better be really cheap to design and fab them.”

What this means for the semiconductor industry will take time to be known. What application semiconductors will be used for, if new materials can be found, or what advances will be seen in quantum or neuromorphic computing are only guesses as yet. So the end of Moore’s Law is not settled yet.

Source: Nature

Leave a Comment

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.