Today, cities are powered around the clock. Businesses and organisations across industry sectors remain online 24 hours a day, seven days a week. With industries increasingly turning to smart technologies to improve work efficiency, society has become reliant on computing infrastructure and the ability to remain connected at all times.
However, powering such infrastructure comes at a price, and as our hunger for electricity grows, so too has our awareness of the impact electricity consumption has on our climate. In 2018, it was estimated that around 33% of the total carbon dioxide emissions in the US were associated with the generation of electricity.
The desire for sustainable living, and in particular, ‘energy efficiency’, is a goal not only for many businesses, but also for individuals that spend their days connected and plugged in.
While computer technology has integrated various energy saving features over the years, from the use of LEDs in display panels, to the inclusion of liquid cooling systems within the processing units, the core of the problem in achieving true energy efficiency has not changed. Our computing hardware will only function if it continues to draw current from the power grid.
Today, there is a long way to go to make our computing infrastructure sustainable — in fact, it is estimated that a single data centre, such as those used by Google or Amazon to store and extract information for their internet search or purchasing services, consumes enough electricity to power 180,000 homes.
Now, one researcher at NUS is hoping to create truly energy efficient devices by combining materials that, when placed together, establish an interface layer where the spin-state of electrons can be switched using a low-voltage electric field alone.
Associate Professor Ariando from NUS Physics, and the NUS Nanoscience and Nanotechnology Initiative, has dedicated his career to understanding oxide materials with unconventional electronic, optical, magnetic and thermal properties.
Whereas traditional computing technology controls and contains information using conventional electronics, where the charge of electrons flowing through a circuit determines the information state, more recent technologies exploit the electron ‘spin’ state to store information.
The ‘spin’-state refers to a quantum property of electrons that sees their alignment, which may be either ‘up’ or ‘down’, generate a magnetic field that serves as a source of measurable and controllable binary data.
The numerous benefits of spin devices include the fact that they are non-volatile, meaning they can retain information even when power is lost. They also function with faster data transfer speeds and with reduced power consumption.
Despite these benefits, the fact that electrical current is required to control spin alignment within magnetic layers limits the true impact of today’s spin-based technologies to reduce our growing reliance on electricity.
As suggested by Assoc Prof Ariando, the concept of controlling spin-based devices without electrical current is achievable, however it will require materials with non-traditional properties. Despite various efforts around the world to synthesise such new materials through chemical processes and composites, they remain non-existent.
To overcome this, Assoc Prof Ariando plans to use his expertise to explore properties that are created at the interface of two otherwise non-magnetic and non-conductive materials. That is, he hopes to artificially engineer new materials by stacking different layers of atoms on top of each other at will, much like playing with Lego at the atomic level.
“Many materials establish unique, and often very useful, properties when they are layered together. These properties often exist at the interface of materials that are otherwise non-functional on their own. It is like taking two slices of bread, putting them together, and suddenly finding a filling has appeared in the middle.” Explained Assoc Prof Ariando.
Using this approach, completely new materials which do not exist in nature, are possible and Assoc Prof Ariando anticipates creating new materials where their spins can be controlled through the application of voltage instead of current.
“Spin-based devices already have huge potential in terms of computing speeds and energy efficiency. If we can remove the need for electrical current in our future computing infrastructure it will translate to savings both economically, and to the environment.” Assoc Prof Ariando continued.
It stands to reason that new materials and devices will indeed be required to meet future computing demands, especially if the new era of smart-tech is to be sustainable. Already, data volumes are surging, and are requiring larger, faster and more stable storage methods. Likewise, the computing resources required to process such volumes of data, in real-time, must continue to evolve to meet this demand.
It is therefore in the pursuit of innovation, where future smart devices could operate using materials and systems that do not require electrical current to function, that true energy efficiency will be found — and this may be just what the world needs right now.
