In fact, according to recent data, if we imagine all data centres forming a single country, this would be the 5^th on the list of countries which consume more energy in the world. And this is expected to get even worse in the coming years, since we –humans- are using more technology every day.

So, in order to tackle this very serious threat, it is of vital importance to develop greener technologies based on radically new, different mechanisms.

Many scientists and engineers worldwide are investigating different approaches. In my case, I use nanotechnology and magnetism. Nanotechnology is the area of technology dealing with systems at the nanoscale, which is normally defined to be between 1 and 100 nm. In order to have an intuitive idea of what these dimensions really mean, let’s look at this ruler. Please focus your attention on the 1 mm scale. Imagine now that you divide this length by 1,000; that length is 1 micron: this distance would be of the same scale as some cells and bacteria. Now divide that distance 1,000 times more: that very small length is 1 nm, which would correspond to about the size of 10 atoms.

Silicon transistors forming integrated circuits in electronic devices are already nanometric, where each bit of information is stored as an electrical charge on a given area of the memory. Normally, a continuous input of energy is needed for this bit to remain in memory and not get erased over time. Also, every time there is a computation, bits are transferred from the memory to the processing unit (CPU), travelling back to memory after the computation. This data transfer between memory and CPU also requires a significant amount of energy.

As you can see, the current computing architecture could be more efficient. A great improvement would result if bits were stable in memory without needing a continuous supply of energy, and if memory and CPU were merged together. This means creating a new computing architecture, a new paradigm for computation.

This is why we investigate the possibility of creating new chips based on networks of magnetic nanowires. A magnetic nanowire is a wire with a diameter at the nanoscale, and a length of a few microns, and has outstanding properties for green computation: Due to its elongated shape, it can store many bits along its length, which can move as tiny confined packages when applying small pulses of current through it. Because it is made of magnetic materials, data is stored without needing a continuous supply of energy. And it is not only great for storage and transmission of data, but it can also perform computations: this is achieved simply by playing with the shape of the wire, and by connecting it to other wires, creating particular geometries.

Many scientists have been studying the properties of magnetic nanowires during the last decade, discovering fascinating new effects. The great challenge now in order to make this technology commercially feasible is to increase the storage capacity of integrated circuits based on them. And this requires an additional radical change, which is creating three-dimensional chips for the first time. Because all integrated circuits that exist and have ever been fabricated are two-dimensional; formed by a plane of transistors and interconnectors, data is confined just to the plane of the 2D chip. This is not casual, and is due to a fundamental manufacturing bottleneck: current nanofabrication methods are able to create complex functional 2D (but not 3D) nanostructures.

In order to overcome this limitation, my group investigates new 3D nanofabrication techniques based on the use of electron microscopes and gases. We’re doing this in close collaboration with other groups. Electron microscopes are one of the most versatile and powerful tools in nanotechnology. By employing electrons instead of conventional light, they are capable of imaging structures at the nanoscale in great detail. In our case, we don’t use these microscopes just for imaging, but also as 3D nano-printers. The images shown below correspond to nanostructures that have been created with this technique, and have features about 1,000 times smaller than what the best 3D printer on the market today can do. This 1-step nanofabrication technique is conceptually much easier than conventional manufacturing methods used by the microelectronics industry, which require tens of steps. But it is not yet suitable for massive mass-production; this will be possible once we find ways to make the technique cheaper and faster.

This one-step nanofabrication technique is conceptually much easier than conventional manufacturing methods used by the microelectronics industry, which require tens of steps. But it is not yet suitable for massive mass-production; this will be possible once we find ways to make the technique cheaper and faster.

In my lab, we’re exploring if the 3D magnetic nanowires manufactured this way have the properties needed for computing applications. We still need to solve key challenges. For example, how to make the wires of better materials, since otherwise we cannot have smooth flow of bits through them. Also, most nanotechnology methods are designed to detect and image 2D objects. We are now creating new ways to resolve the magnetic properties at the nanoscale in 3D.

I find very exciting that we’re now closer to create electronic circuits which will be greener and that will have novel functionalities. This will make computing devices more independent of energy sources, becoming much less of a concern for the global warming of the planet. In order to achieve this, I would like to plead for a society that is aware of the challenges of the future. A society that supports research to advance and overcome these challenges. And that makes a responsible use of technology, so that coming generations can continue enjoying our now, technological world.