Outreach

Neuromorphic Computing: A Brief Explanation

Have you ever thought about why we can not perform brain tasks on our computers? Well of course I don’t mean a simple cat/dog recognition, or calculus (Computers have been specifically designed to do a very limited number of brain functions extremely well – even better than humans), but something bigger like analyzing new and unfamiliar situations.

To answer this question, let’s first see how conventional computers work:

Simply explained, there are two main units: a processing unit to process and analyze the data and a memory unit to store them. These two blocks are separated from each other and every time that a task must be done, the data should go back and forth between these two units. This architecture is known as Von Neumann architecture [1].

Fig1: Von Neumann architecture: in a conventional computing system, when an operation f is performed on data D, D has to be moved into a processing unit, leading to significant costs in latency and energy [2].

As you’ve already found out, there are two issues with this architecture that makes it almost impossible to do heavy tasks with:

  1. Energy consumption, as the blocks are “separated” and lots of Joule heating can happen in between.
  2. Not fast enough, due to the time required for the data to go back and forth.

This is also known as the Von Neumann bottleneck [3]. In other words, the architecture causes a limitation on the throughput, and this intensive data exchange is a problem. To find an alternative for it, it’s best to take a look at our brain and try to build something to emulate it, because not only is it the fastest computer available, but it is super energy efficient.

The brain is made up of a very dense network of interconnected neurons, which are responsible for all the data processing happening there. Each neuron has three parts: Soma (some call it neuron as well) which is the cell body and is responsible for the chemical processing of the neuron, Synapse which is like the memory unit and determines the strength of the connections to other neurons, and Axon which is like the wire connecting one neuron to the other.

Fig2: Neural networks in biology and computing [4].

Neurons communicate with voltage signals (spikes) generated by the ions and the chemicals inside our brains. There have been many models presented on how they work, but here will be discussed the simplest (and probably the most useful) one: The leaky integrate and fire model [5].

Fig3: leaky integrate and fire model, Incoming pulses excite the biological neuron; if the excitation reaches a threshold, the neuron fires an outcoming spike, and if not, it relaxes back to the resting potential [6].

As it was said earlier, neurons communicate with spikes which can change the potential of the soma. If a spike from a previous neuron arrives at a neuron after it, the potential of the soma increases. However, this is temporary, meaning that if no other spikes arrive afterward, the potential of the soma will reach the relaxed level again (leakage). On the other hand, if a train of spikes arrives at the neuron, they can accumulate (integrate) and if the potential reaches a threshold potential, the neuron itself will generate a spike (fire). After firing, the potential will again reach the relaxed level.  

Apparently, the connections between all the neurons are not the same, and the differences are in the synapses. The form and combination of the synapses change in time depending on how active or inactive those two neurons were. The more they communicate, the stronger and better their connection, and this is called “synaptic plasticity”. (This is why learning something new is so hard because the connections between the neurons need time and practice to get better!). For more investigation into the fascinating world of the brain, this book is recommended: Neuronal Dynamics: From Single Neurons to Networks and Models of Cognition [7].

Now, it’s time to get back to the Von Neumann bottleneck. With inspiration from the brain, it can be seen that it’s better to place the memory unit in the vicinity, or even inside, the processing unit (just like the soma and the synapses which are really close), this way so much time and energy can be saved. It is also obvious that the processing units are better to be nonlinear as in the brain, and the memory unit should be able to be changed or manipulated to mimic the synaptic plasticity. We know how different parts should behave in order to have a computer to at least function like the brain, but the big question is: What hardwares should be used? What kind of devices act like a neuron, or a synapse? And even if we find them, are we able to place them close to each other to overcome the Von Neumann bottleneck?

These are the questions that Neuromorphic computing tries to answer. In other words, it is an attempt to build new hardware to be able to do computing like our brain. Some of the most promising candidates here are the spin-orbit devices as they are super-fast, energy-efficient, and more importantly, nonlinear [8][9]. I will talk about them and their major role in this field more in detail in the second part of my post soon!

Please don’t hesitate to ask questions: mahak@chalmers.se

References:

1. Von Neumann, J. Papers of John von Neumann on computers and computer theory. United States: N. p., 1986. Web.

2. Sebastian, A., Le Gallo, M., Khaddam-Aljameh, R. et al. Memory devices and applications for in-memory computing. Nat. Nanotechnol. 15, 529–544 (2020).

3. John Backus. 1978. Can programming be liberated from the von Neumann style? a functional style and its algebra of programs. Commun. ACM 21, 8 (Aug. 1978), 613–641.

4. Bains, S. The business of building brains. Nat Electron 3, 348–351 (2020).

5. Brunel, N., van Rossum, M.C.W. Lapicque’s 1907 paper: from frogs to integrate-and-fire. Biol Cybern 97, 337–339 (2007).

6. Kurenkov, A., DuttaGupta, S., Zhang, C., Fukami, S., Horio, Y., Ohno, H., Artificial Neuron and Synapse Realized in an Antiferromagnet/Ferromagnet Heterostructure Using Dynamics of Spin–Orbit Torque Switching. Adv. Mater. 2019, 31, 1900636.

7. Gerstner, W., Kistler, W., Naud, R., & Paninski, L. (2014). Neuronal Dynamics: From Single Neurons to Networks and Models of Cognition. Cambridge: Cambridge University Press.

8. Grollier, J., Querlioz, D., Camsari, K.Y. et al. Neuromorphic spintronics. Nat Electron 3, 360–370 (2020).

9. Zahedinejad, M., Fulara, H., Khymyn, R. et al. Memristive control of mutual spin Hall nano-oscillator synchronization for neuromorphic computing. Nat. Mater. 21, 81–87 (2022).

Ferroelectric materials and applications

Ferroelectrics are a category of material that in absence of an external applied voltage they still show a remanent polarization. The reason can be found in their atomic structure. For example the family of perovskites are ferroelectric material due to their specific crystal arrangement. Perovskite all share this similar ABX3 structure, where usually the X is oxygen and the A and B represent two different metals.

Fig.1 Phases of KNbO3 (potassium niobate) at different temperatures. It shows some structures where is possible to have a remanent polarization due to non-centro symmetry of the Niobium atom (in green) inside the cubic structure. When the Niobium atom is exactly at the center (in this case above 708 K) the material is not anymore ferroelectric [1].

One of most famous and studied is lead zirconate (PbZrxTi1-xO3) commonly called PZT. One of the biggest issues with this material is the toxicity due to the presence of the lead. For this reason, researchers focused on finding a material with similar characteristics. So many of them came out like barium titanate (BaTiO3), know as BTO or strontium titanate (SrTiO3) known as STO. In the picture here above another example of a lead-free perovskite material: potassium niobate (KNO3).

The remanent polarization properties is given by the presence of an atom inside this cubic-like structure that is not exactly at the center but slightly shifted. This non centro-symmetric structure give rise to a non-compensated positive charge of the body atom (Ti or Nb for example). The ferroelectric properties then are just given by the presence of the non-compensated charge when all the external voltages are removed. The temperature has an important role since for every material, for a given energy the structure tend to become a symmetric body centered cubic structure, hence there is a critical temperature after which the ferroelectric materials become paraelectrics. In this state they still react non-linearly to an external applied field but they do not show a remanent effect in absence of it.

Fig.2 Behaviour of a dielectric, a paraelectric and a ferroelectric material under an applied external electric field [2].

The polarization bistability for a zero external applied field is the key feature that makes ferroelectrics good candidates for memory applications. In recent year, many studies focused on integrating the ferroelectric materials in order to create new memories, more competitive from an energy computation point of view or overall faster writing and reading speed as FE-Fet [3], FE-RAM [4] or MESO [5] and FESO [6]. If for the first two the ferroelectric properties are aimed to improve the properties of already existing devices, like transistors or non-volatile random access memories; for the MESO and FESO case the aim is more ambitious. The idea is to develop a new logic based on spin controlled by ferrolectric non-volatility.

Recently, others materials showed to have ferroelectricity properties like Germanium telluride (GeTe), Indium arsenide (InAs) and many others. These materials show a simpler combination of only two atoms and that are not insulating like perovskites (sually they are metalic or semiconducting).

The bigger advantage is the possibility to pattern them in order to produce nanodevices, given by an higher durability when subjected to nanofabrication steps like etching. This one tend to destroy the crystal structure and hence the properties of the insulating perovskites . As a consequence, these new materials bring the ferroelectric-based devices a step closer to mass production and adoption.

If we take into account the case of germanium telluride, the ferroelectricity comes from the unusual bonds between germanium and tellurium layers. They tend to form a stronger bond with a neighbour layer with respect to the other forming a bilayer structure that is not symmetric (see pictures below). Similarly under an applied electric field the structure reorganize, causing the polarization to change sign (if the field is strong enough). It can be also seen as the germanium in the center of the cell moving along the larger diagonal of the deformed cubic cell (also called rhombohedral cell, left picture).

Fig. 3  Left: Cell structure of Germanium telluride (yellow Germanium, blue tellurium). Right: Switching mechanism: a) stable configuration of layer of Germanium (yellow dots) bonded to Tellurium ones (in red). b) when an electric field is applied, an unstable state appear where Germanium is bond to both top and bottom Tellurium atoms. c) Final state in which the Germanium atoms will be bond to the Tellurium atoms in the upper level with respect to the initial state [7]

So in a similar way to perovskites germanium telluride is know to show a remanent polarization at room temperature that can be controlled by an external applied electric field.

I hope you enjoyed this small talk on ferroelectrics, I will write in future a part 2 to explain the relation between ferroelectricity and spin-logic based devices. For further information you can email me at: salvatore.teresi@cea.fr

References

[1] P. Hirel et al., Phys. Rev. B 92 (2016) 214101.

[2] http://faculty-science.blogspot.com/2010/11/ferroelectricity.html

[3] Stefan Ferdinand Müller (2016). Development of HfO2-Based Ferroelectric Memories for Future CMOS Technology Nodes. ISBN 9783739248943.

[4] Dudley A. Buck, “Ferroelectrics for Digital Information Storage and Switching.” Report R-212, MIT, June 1952.

[5] Manipatruni, S., Nikonov, D.E., Lin, CC. et al. Scalable energy-efficient magnetoelectric spin–orbit logic. Nature 565, 35–42 (2019). https://doi.org/10.1038/s41586-018-0770-2

[6] Noël, P., Trier, F., Vicente Arche, L.M. et al. Non-volatile electric control of spin–charge conversion in a SrTiO3 Rashba system. Nature 580, 483–486 (2020). [7] A. V. Kolobov, D. J. Kim, A. Giussani, P. Fons, J. Tominaga, R. Calarco, and A. Gruverman, Ferroelectric switching in epitaxial GeTe films, APL Materials 2, 066101 (2014).

[7] A. V. Kolobov, D. J. Kim, A. Giussani, P. Fons, J. Tominaga, R. Calarco, and A. Gruverman, Ferroelectric switching in epitaxial GeTe films, APL Materials 2, 066101 (2014).

Antiferromagnetism

Hello readers! This first post is about antiferromagnets. I talk about the general concepts in antiferromagnetism and then later I dive into some interesting complex antiferromagnetic states at the atomic scale! I hope you have a good read! Please feel free to reach me out at my email for any further questions! (vsaxena@physnet.uni-hamburg.de)

Click here to check out the post And stay tuned for my next one! I will come back some more interesting magnetism!

Diseño y desarrollo web Triplevdoble