Recently, there has been impressive progress in the field of artificial intelligence. A striking example is Alphago, an algorithm developed by Google, that defeated the world champion Lee Sedol at the game of Go. However, in terms of power consumption, the brain remains the absolute winner, by four orders of magnitudes. Indeed, today, brain inspired algorithms are running on our current sequential computers, which have a very different architecture than the brain. If we want to build smart chips capable of cognitive tasks with a low power consumption, we need to fabricate on silicon huge parallel networks of artificial synapses and neurons, bringing memory close to processing. The aim of the presented work is to deliver a new breed of bio-inspired magnetic devices for pattern recognition. Their functionality is based on the magnetic reversal properties of an artificial spin ice in a Kagome geometry for which the magnetic switching occurs by avalanches.
In physics, frustration appears in a system when it is impossible to minimise all pairwise interactions simultaneously. Frustration exists in some particular rare-earth based compounds, such as spin ices . Their internal frustration gives rise to unusual properties, like a residual entropy at low temperature or the presence of monopole-like excitations . However, experimental techniques are unable to probe each spin individually in these compounds.
In 2006, Wang and coworkers opened a new way for studying magnetically frustrated spin systems . Using e-beam lithography, one can make arrays of nanomagnets with the desired design. The state of each nanomagnet can then be probed individually in real space at room temperature using magnetic imaging (eg. Magnetic Force Microscopy). In this context, the square geometry received a considerable interest, since it is closely related to condensed matter spin ice compounds. But for geometrical reasons, this system orders instead of showing a disordered low energy manifold
In this contribution, we explain how to bring back the massive ground state degeneracy in the square array of nanomagnets. We present the first experimental evidence of a Coulomb phase in this system . We also report the presence of magnetic monopoles defects within the Coulomb phase. This study makes a new step toward a direct study of the dynamic of monopoles excitations (e.g. creation, annihilation or diffusion processes).
 M.J. Harris et al., Phys. Rev. Lett. 79, 2554 (1997).
 C. Castelnovo et al., Nature 451, 7174 (2008).
 R.F. Wang et al., Nature 439, 303 (2006).
 Y. Perrin et al., Nature 540, 410 (2016).
Complex architectures of nanostructures are currently routinely elaborated using bottom-up or nanofabrication processes. This technological capability allows scientists to engineer materials with properties that do not exist in nature, but also to manufacture model systems to explore fundamental issues in condensed matter physics. Two-dimensional frustrated arrays of magnetic nanostructures are one class of systems for which theoretical predictions can now be tested experimentally.
In particular, magnetic imaging techniques offer the appealing opportunity to observe a wide range of phenomena within the concept of lab-on-a-chip. For example, several exotic magnetic phases have been discovered in artificial frustrated spin systems. Besides, these systems allow the study of classical analogues of magnetic monopoles. These recent results have stimulated new research activities motivated by the quest for magnetic monopoles in condensed matter physics.
In this contribution, we'll show that the micromagnetic properties of the elements constituting artificial frustrated arrays of nanomagnets introduce the concept of chiral monopoles. Injecting and manipulating experimentally the chirality of a magnetic monopole provide a new degree of freedom in the system. This offers the opportunity to control their motion under an external magnetic field, thus allowing to envision applications in magnetronics.
CMOS technology allows a femto Joule energy dissipation per logic operation, if operated at optimal frequency and voltage. However, this value remains orders of magnitude above the theoretical limit predicted by Lan-dauer. In this work, we present a new paradigm for low power computation, based on variable capacitors. Such components can be implemented with existing MEMS technologies. We show how a smooth capacitance modu-lation allows an energy-eﬃcient transfer of information through the circuit. By removing electrical contacts, our method limits the current leakages and the associated energy loss. Therefore, capacitive logic must be able to achieve extremely low power dissipation when driven adiabatically. Contactless capacitive logic also promises a better reliability than systems based on MEMS nanorelays.
Artificial arrays of interacting magnetic elements provide an uncharted arena in which the physics of magnetic frustration and magnetic monopoles can be observed in real space and in real time. These systems offer the formidable opportunity to investigate a wide range of collective magnetic phenomena with a lab-on-chip approach and to explore various theoretical predictions from spin models. Here, we study artificial square ice systems numerically and use micromagnetic simulations to understand how the geometrical parameters of the individual magnetic elements affect the energy levels of an isolated square vertex. More specifically, we address the question of whether the celebrated square ice model could be made relevant for artificial square ice systems. Our work reveals that tuning the geometry alone should not allow the experimental realization of the square ice model when using nanomagnets coupled through the magnetostatic interaction. However, low-aspect ratios combined with small gaps separating neighboring magnetic elements of moderated thickness might permit approaching the ideal case where the degeneracy of the ice rule states is recovered.
Complex architectures of nanostructures are routinely elaborated using bottom-up or nanofabrication processes. This technological capability allows scientists to engineer materials with properties that do not exist in nature, but also to manufacture model systems to explore fundamental issues in condensed matter physics. Two-dimensional frustrated arrays of magnetic nanostructures are one class of systems for which theoretical predictions can be tested experimentally. These systems have been the subject of intense research in the last few years and allowed the investigation of a rich physics and fascinating phenomena, such as the exploration of the extensively degenerate ground-state manifolds of spin ice systems, the evidence of new magnetic phases in purely two-dimensional lattices, and the observation of pseudoexcitations involving classical analogues of magnetic monopoles. We show here, experimentally and theoretically, that simple magnetic geometries can lead to unconventional, non-collinear spin textures. For example, kagome arrays of inplane magnetized nano-islands do not show magnetic order. Instead, these systems are characterized by spin textures with intriguing properties, such as chirality, coexistence of magnetic order and disorder, and charge crystallization. Magnetic frustration effects in lithographically patterned kagome arrays of nanomagnets with out-of-plane magnetization also lead to an unusal, and still unknown, magnetic ground state manifold. Besides the influence of the lattice geometry, the micromagnetic nature of the elements constituting the arrays introduce the concept of chiral magnetic monopoles, bringing additional complexity into the physics of artificial frustrated spin systems.