What NYU and IBM’s Research May Tell us About Spintronics in Computing

3 weeks ago by Kristijan Nelkovski

In early 2020, NYU (New York University) and the magnetic memory development team at IBM Research revealed a potential method to improve data storage using the motion of electrons in magnetic materials. In view of the partnership’s breakthroughs, we discuss the clear potential of spintronics in modern computing.

A Recap on Spintronics

Electrons in physics are characterised by three fundamental properties: the ‘mass’, ‘charge’, and ‘spin’. As previously covered on Electronics Point, much of modern electronics (such as processors and data storage devices) stem from the manipulation of the charge of electrons in order to execute computations and store data. Semiconductor electronics works by defining a positive electrical charge as a logic 1 and a negative electrical charge as a logic 0.

In addition to the electron charge, the quantum mechanical property of the electron, i.e. the spin, can also be utilised to represent logic states. This is as it can take two values (spin up and spin down), and this offers an additional degree of freedom for the electron. This degree of freedom is separate from the charge, and it opens up new possibilities for more efficient data storage.

Fast computing concept. Pictured: a graphic of a flying rocket positioned next to a laptop to represent the high-speed computations facilitated by spintronics. Image Credit: Bigstock.


More Information on Spintronics

The field that studies the above phenomena is called spintronics (spin electronics), which describes devices that operate by using both the spin of the electron as well as its electrical charge. This research carried out by NYU and IBM’s partnership (which you can also read about on Electronics Point’s sister site, All About Circuits) focuses on an important advancement in existing spintronics research on data storage.

As Andrew D Kent—a professor at NYU and one of the physicists behind the research—points out, the goal of spintronics researchers is in finding the ability to control the spin of an electron in various materials. Kent’s research demonstrates a new way of setting the spin direction of an electron in a conducting material.

Jonathan Sun, a senior co-author from IBM Research, explains that their analysis opens up a potential new way in which torque can be applied to an electron within the magnetic layer of a given material. This breakthrough allows for data storage devices to use less energy while storing more data, particularly by utilising the spin of the electrons within the material. This achieves the storage of information in a more compact and efficient way.

Transforming data like this into a much more compact form makes for faster read and write speeds, and it also facilitates an easier way to store and manipulate information overall.


The Use of the Hall Effect in Spintronics

The basis for spintronics is the Hall Effect: a physical phenomenon that is exhibited when a perpendicular magnetic field is introduced to an electric conductor or semiconductor that has a unidirectional current flowing through it. When this occurs, a voltage difference is produced within the conductor (or semiconductor) transverse to its current, flowing towards the magnetic field (or otherwise, depending on the magnet’s polarity, against it).

Analogous to the Hall Effect is the Spin Hall Effect. This is where spin accumulation occurs in a material that carries electrical current. In the past, this effect has been shown in non-magnetic heavy metals, which exhibit spin polarisation on the surface of the conductor.

The Implementation of a Ferromagnetic Conductor

One of the problems with using the Spin Hall Effect for spintronic data storage devices is that the spin polarisation of the electrons in these materials always happens parallel to the material surface. Because of this, the NYU and IBM scientists utilised the planar Hall effect in a ferromagnetic conductor in order to guide the spin polarisation of the electrons.

The ferromagnetic conductor used in this research allows for current flowing through it to produce a spin polarisation that moves in the direction guided by its magnetic moment. This differs from the Spin Hall Effect because, in this case, the spin of the electron can be adjusted in any direction—and not just parallel to the magnetic surface. This opens up a gateway in spintronic data storing and processing with much better characteristics, particularly in terms of read and write speeds, computing capabilities, as well as both the volatility and size of the disk itself.

‘Opening up’ the potential of future data storage. Pictured: the inside of a traditional 2.5-inch hard drive is revealed by opening up the device’s front protective cover. Image Credit: Pixabay.


What Spintronics Means for the Future of Data Storage and Manipulation

Modern HDD (hard disk drive) technology doesn’t have the ability to scale down due to the size restrictions of the disc’s read and write elements, as well as the fundamental physics limitation in traditional electronics. Modern RAM (random-access memory) technology also faces the problem of physical limitations, all while being a volatile type of memory, which means that it must be constantly powered in order to store data. 

Ultimately, the New York University-IBM research has established the potential of spintronic devices that can remove the above-mentioned bottlenecks and other current spintronics limitations. IBM has previously worked with multiple universities and organisations in this area, as they have explored the properties of spintronics and figured out methods to develop advanced technology based on these properties.

By using this new discovery in early 2020, what IBM and NYU did was collectively establish the potential of a new class of spintronic data devices that may have a smaller size and a significant increase in functionality—allowing not only the improvement of current technology but also future advancements in the area of quantum computing, wherein the quantum qubits can occupy both the spin up and spin down states simultaneously. Matters such as this will contribute to enormous increases in computational power.