Silicon Chip Replacement
Silicon chips are the backbone of the electronics industry. They are made of a very strong, transparent material that conducts electricity and heat well.
But the technology is getting old and we need to find a replacement. Otherwise, Moore’s Law will end and technological progress will stop.
Fortunately, emerging technologies like graphene, quantum computing, and Gallium Nitride show promise as potential silicon chip replacements.
Quantum computing
When people hear the word “quantum computing”, they often think of machines that will revolutionise everything from drug discovery to financial transactions. They might also speed up machine learning, improve stock market forecasting and even solve climate change. These computers exploit the strange properties of atoms to solve complex problems, and they can perform calculations much faster than classical supercomputers.
Unlike the billions of transistors that handle data on a modern computer, quantum processors work in two states: either on or off. These are known as qubits, and the more they have, the more powerful they become.
Engineers have been working to develop solid, crystalline structures that could be used as the building blocks for future quantum computers. These so-called atomic crystals could be used for the qubits that will power future computers, and they might provide better performance than existing silicon technology. These materials would be much cheaper to produce, too. But these breakthroughs will take years to translate into practical applications.
Graphene
Since its discovery in 2004, graphene has sparked interest across industries. It is the thinnest material ever made, with a single atom-thick layer, and it is 100 times stronger than steel. It is also a great conductor of heat and electricity. Graphene can be used to make strong and lightweight materials for batteries, transparent touchscreens, and other electronic devices. It can also be used to reinforce plastics and paints, or to create conductive inks for use on chips.
It may not replace silicon in the near future, but it has significant potential. It could be used to improve transistor lifetime and enable smaller feature sizes for CPUs and memory. In addition, it can be used to develop new materials with unique electrical properties.
Although the development of graphene is still at an early stage, it is attracting increasing interest from semiconductor companies. However, executives must take a broader view of the market to extract the most value from the emerging technology. This requires a different mindset that focuses on connecting the dots between seemingly disparate developments.
Carbon nanotubes
Carbon nanotubes, which are tiny rolled-up pieces of the 2D wonder material graphene, could make transistors that would be five times more energy efficient than silicon chips. But getting them to work isn’t easy. They need to be deposited uniformly, exhibit consistent electrical characteristics, and be close enough together to carry current. But engineers at the University of Wisconsin-Madison have found a way to get those little tubes lined up just right.
The key is removing contaminants. When carbon nanotubes are grown, they tend to be a mixture of metallic and semiconducting. That’s a big problem in a chip, where the electrical properties depend on the (n,m) type of each tube.
To prevent contamination, the team has developed a new method of preprocessing the wafers before they are incubated with carbon nanotubes. The wafers are dipped in a solution that shakes away bundles of carbon, leaving only the single tubes behind. These can be used to form the sources, drains, and gates of a transistor.
Nanomagnetic logic
In contrast to silicon transistors, which use electric current, nanomagnetic logic uses magnetic fields to process binary code. This allows for more complex circuits to be produced without consuming as much power. In addition, nanomagnetic logic can be fabricated with the same fabrication processes as CMOS chips.
A key challenge for future ICs is to reduce energy consumption, and nanomagnetic logic may be the solution. The technology uses magnetic field-based logic, which is similar to quantum computing but consumes a fraction of the power. It works by using the interaction of the north and south poles of each magnet to create a binary code.
Nanomagnetic logic chains are a promising computational architecture that promises ultra-low energy dissipation per operation. They are based on pulsed nanosecond on-chip magnetic fields8 that drive the magnetization of all islands in a chain to saturation along their magnetic hard axis. The resulting near-null state of adjacent magnets allows for nearest-neighbour dipolar field coupling, which can be used to form spatial logical inverters.