In an age where everything is becoming increasingly computerized, computer circuits powered practically everything. Modern life is dependent on semiconductor chips and transistors on silicon-based integrated circuits, which have the power to switch electronic signals on and off. Cheap element silicon is prevalent in these circuits since it can be used to control the flow of electricity, as it can insulate and semiconduct electricity.
Until recently, microscopic transistors squeezed onto silicon chips have halved in size each year. That reduction in size has paved the way for modern technology, yet people like Jamie Carter of TechRadar believe that silicon chips may soon outlive their usefulness. The modern digital age that semiconductors helped create has lead to the advent of the Internet of Things (IoT), Artificial Intelligence (AI), autonomous vehicles, and 5G phones.
In 1956, Gordon Moore the co-founder of Intel, observed that the number of transistors on a one-inch computer chip would double each year, while costs would halve. Now the transistors are doubling every 18 months, and that period might get longer. Carter believes that Moore’s Law is not a set-in-stone truth as much as it is “an observation by someone who worked for a chip-maker.” He went on to say that “the increased timescales mean intensive computer applications of the future could be under threat.”
“Silicon is reaching the limit of its performance in a growing number of applications that require increased speed, reduced latency, and light detection,” says Stephen Doran, CEO of the UK’s Compound Semiconductor Applications Catapult. However, he believes that it is still too early to think of a successor to silicon. “That suggests silicon will be completely replaced, which is unlikely to happen anytime soon, and may well never happen.”
While Moore’s Law might fade away, experts think that silicon will still play a prominent role well into the foreseeable future. “There is still potential in a Moore’s Law-style performance escalation until at least 2025,” says David Harold, Vice President of Marketing Communication at Imagination Technologies. “Silicon will dominate the chip market until the 2040s.”
Craig Hampel, Chief Scientist for the Memory and Interface Division at Rambus also believes that some perspective on the silicon transistor issue is needed. “Moore’s Law specifically refers to the performance of integrated circuits made from semiconductors, and only captures the last 50-plus years of computation.”
Hampel believes that one has to look at the bigger picture to understand humanity’s need for computation, which “reaches back to the abacus, mechanical calculators, and vacuum tubes, and will likely extend well beyond semiconductors to include superconductors and quantum mechanics.”
The major issue is that computers need to be more powerful and more agile. Harold believes that that the demand for future systems to be more brain-like may “create a revolutionary second era for computing.”
More Power for Less
Researchers are looking for new ways of achieving higher performance in computers while using less power. “Cold operation of data centers of supercomputers can have significant performance, power, and cost advantage,” says Hampel.
Cold computation reduces some of the challenges of extending Moore’s Law. The natural operating temperature for cold computing machines is that of liquid nitrogen at 77K (-270C). “Nitrogen is abundant in the atmosphere, relatively inexpensive to capture in liquid form, and an efficient cooling medium,” says Hampel. “We hope to get perhaps four to ten additional years of scaling in memory performance and power.”
Another solution being examined are next-gen semiconductors made from two or more elements, making them faster and more efficient than silicon. These chips are already being utilized in future 5G and 6G phones. “Compound semiconductors combine two or more elements from the periodic table, for example gallium and nitrogen, to form gallium nitride,” says Doran. Their ability to outperform silicon makes applications like 5G and autonomous vehicles possible.
There is also quantum computing. IBM, Google, Intel, and others are in a race to create quantum computers with enormous processing power that would dwarf silicon transistors, using quantum bits. However, the problem is that quantum mechanics is still very much an uncharted, nebulous field, where quantum physicists and computer architects need to make many breakthroughs before the full potential of the technology can be realized. There’s a simple test that some in the quantum computing community think needs to be achieved before an actual quantum computer can exist: “quantum supremacy.”
“It means simply showing that a quantum machine is better at a specific task than a conventional semiconductor processor on the path of Moore’s Law,” says Hampel. So far, quantum computing is still within the realm of the theoretical.
The Bottom Line
The only constant force in the world is change. Moore’s Law has been a reliable model for predicting how silicon-based semiconductors have evolved over the last couple of decades. However, Moore’s Law has grown less reliable over time as the period over which semiconductors upgrade has been extended. Nevertheless, Moore’s Law is less of a set-in-stone law rather than an observation, and that semiconductors still have a place in the world of computing.
Nevertheless, alternative solutions are being explored, ranging from cold computing, to using alternative compounds to silicon in semiconductors, to exploring quantum computing. All these solutions are able to produce more computing power than their silicon semiconductor counterparts. It seems there will always been a need for coming up with exponential improvements to computing. In the words of Hampel, “while Moore’s Law will end, the secular and lasting trend of exponential computing capacity will likely not.”