Massimiliano Di Ventra

University of California, San Diego | UCSD · Department of Physics

The "criticality hypothesis", based on observed scale-free correlations in neural activity, suggests that the animal brain operates at a critical point of transition between two phases. However, what these phases are, or whether such a hypothesis is at all valid, is still debated. Here, using a cortical dynamics model based on the work of Wilson an...

In the pursuit of scalable and energy-efficient neuromorphic devices, recent research has unveiled a novel category of spiking oscillators, termed “thermal neuristors.” These devices function via thermal interactions among neighboring vanadium dioxide resistive memories, emulating biological neuronal behavior. Here, we show that the collective dyna...

Time non-locality, or memory, is a non-equilibrium property shared by all physical systems. It means that when a system's state is perturbed, it is still affected by the perturbation at a later time. Here, we show that such a memory effect is sufficient to induce a phase of spatial long-range order (LRO) even if the system's dynamical variables are...

In the “Beyond Moore’s Law” era, with increasing edge intelligence, domain-specific computing embracing unconventional approaches will become increasingly prevalent. At the same time, the adoption of a wide variety of nanotechnologies will offer benefits in energy cost, computational speed, reduced footprint, cyber-resilience and processing prowess...

Neuromorphic systems are typically based on nanoscale electronic devices, but nature relies on ions for energy-efficient information processing. Nanofluidic memristive devices could thus potentially be used to construct electrolytic computers that mimic the brain down to its basic principles of operation. Here we report a nanofluidic device that is...

The Peltier effect, which is the reverse counterpart of the Seebeck effect, has been experimentally observed in nanojunctions. Despite its potential applications in cooling nanoelectronic devices, achieving significant figures of merit and cooling power remain challenging. Here, we propose a novel approach to enable substantial Peltier cooling by l...

In the pursuit of scalable and energy-efficient neuromorphic devices, recent research has unveiled a novel category of spiking oscillators, termed ``thermal neuristors." These devices function via thermal interactions among neighboring vanadium dioxide resistive memories, closely mimicking the behavior of biological neurons. Here, we show that the...

While the complementary metal‐oxide semiconductor (CMOS) technology is the mainstream for the hardware implementation of neural networks, we explore an alternative route based on a new class of spiking oscillators we call “thermal neuristors”, which operate and interact solely via thermal processes. Utilizing the insulator‐to‐metal transition in va...

Digital MemComputing machines (DMMs), which employ nonlinear dynamical systems with memory (time non-locality), have proven to be a robust and scalable unconventional computing approach for solving a wide variety of combinatorial optimization problems. However, most of the research so far has focused on the numerical simulations of the equations of...

Digital memcomputing machines (DMMs) are a new class of computing machines that employ nonquantum dynamical systems with memory to solve combinatorial optimization problems. Here, we show that the time to solution (TTS) of DMMs follows an inverse Gaussian distribution, with the TTS self-averaging with increasing problem size, irrespective of the pr...

While the complementary metal-oxide semiconductor (CMOS) technology is the mainstream for the hardware implementation of neural networks, we explore an alternative route based on a new class of spiking oscillators we call thermal neuristors, which operate and interact solely via thermal processes. Utilizing the insulator-to-metal transition in vana...

While most neuromorphic systems are based on nanoscale electronic devices, nature relies on ions for energy-efficient information processing. Therefore, finding memristive nanofluidic devices is a milestone toward realizing electrolytic computers mimicking the brain down to its basic principles of operations. Here, we present a nanofluidic device d...

Memcomputing is a novel computing paradigm beyond the von-Neumann one. Its digital version is designed for the efficient solution of combinatorial optimization problems, which emerge in various fields of science and technology. Previously, the performance of digital memcomputing machines (DMMs) was demonstrated using software simulations of their o...

Inspired by the advancements in large language models based on transformers, we introduce the transformer quantum state (TQS): a versatile machine learning model for quantum many-body problems. In sharp contrast to Hamiltonian/task specific models, TQS can generate the entire phase diagram, predict field strengths with experimental measurements, an...

MemComputing is a new model of computation that exploits the non-equilibrium property-we call 'memory'-of any physical system to respond to external perturbations by keeping track of how it has reacted at previous times. Its digital, scalable version maps a finite string of symbols into a finite string of symbols. In this paper, I will discuss some...

In this chapter we show that the mathematical definition of memristors and other memory circuit elements (such as memcapacitors and meminductors) is simply a special case of the most general notion of response functions we have introduced in the previous chapter. It will then be clear that (i) memelements are not fundamental circuit elements, and (...

In this chapter we introduce an unambiguous test to experimentally determine whether a given device is an ideal memristor or not. We demonstrate the test by applying it to resistive random-access memory (ReRAM) cells and the so-called “\(\Phi \) memristor.” In both cases, the experimental test clearly shows that, unlike what has been claimed in the...

In this chapter we introduce the notion of memory as a general property common to all physical systems. We emphasize that any physical system subject to suitable perturbations showcases some degree of memory, whether this memory is easy or not to detect experimentally. In the Physics literature, the description of memory effects is provided by the...

In this concluding chapter we summarize our thoughts on the field, its sociology, and where it could go from here. The main lesson we draw is that blind and unquestioned reliance on mathematical definitions to describe physical phenomena can easily lead to wrong conclusions and away from physical reality.

In this chapter we show that ideal memristors (as discussed in the previous Chap. 2) are subject to very strict physical conditions and are unable to protect their memory state against the unavoidable fluctuations, and therefore are susceptible to a stochastic catastrophe. Similar considerations apply to ideal memcapacitors and meminductors. These...

Digital MemComputing machines (DMMs) are a new class of computing machines that employ non-quantum dynamical systems with memory to solve combinatorial optimization problems. Here, we show that the time to solution (TTS) of DMMs follows an inverse Gaussian distribution, with the TTS self-averaging with increasing problem size, irrespective of the p...

In the Beyond Moore Law era, with increasing edge intelligence, domain-specific computing embracing unconventional approaches will become increasingly prevalent. At the same time, the adoption of a wide variety of nanotechnologies will offer benefits in energy cost, computational speed, reduced footprint, cyber-resilience and processing prowess. Th...

A physical system that retrieves information from the past and acts on it appropriately can efficiently solve difficult combinatorial-optimization problems.

Neural networks (NNs) representing quantum states are typically trained using Markov chain Monte Carlo based methods. However, unless specifically designed, such samplers only consist of local moves, making the slow-mixing problem prominent even for extremely simple quantum states. Here, we propose to use mode-assisted training that provides global...

Recent advancements in machine learning have led to the introduction of the transformer, a versatile, task-agnostic architecture with minimal requirements for hand-crafting schemes and features across different tasks. Here, we show that with appropriate modifications, such an architecture is well suited as a multi-purpose model for the solution of...

A simple and unambiguous test has been recently suggested [J. Phys. D: Applied Physics, 52, 01LT01 (2018)] to check experimentally if a resistor with memory is indeed a memristor, namely a resistor whose resistance depends only on the charge that flows through it, or on the history of the voltage across it. However, although such a test would repre...

From the originator of MemComputing comes the very first book on this new computing paradigm that employs time non-locality (memory) to both process and store information. The book discusses the rationale behind MemComputing, its theoretical foundations, and wide-range applicability to combinatorial optimization problems, Machine Learning, and Quan...

From the originator of MemComputing comes the very first book on this new computing paradigm that employs time non-locality (memory) to both process and store information. The book discusses the rationale behind MemComputing, its theoretical foundations, and wide-range applicability to combinatorial optimization problems, Machine Learning, and Quan...

This introductory chapter sets out the theoretical foundations for ‘deliberative accountability’ and describes the three metrics by which to evaluate the quality of deliberative accountability—namely, respect among actors, the extent to which partisan-point scoring replaces reason-giving, and a reciprocal dialogue. The context in which these are ex...

From the originator of MemComputing comes the very first book on this new computing paradigm that employs time non-locality (memory) to both process and store information. The book discusses the rationale behind MemComputing, its theoretical foundations, and wide-range applicability to combinatorial optimization problems, Machine Learning, and Quan...

From the originator of MemComputing comes the very first book on this new computing paradigm that employs time non-locality (memory) to both process and store information. The book discusses the rationale behind MemComputing, its theoretical foundations, and wide-range applicability to combinatorial optimization problems, Machine Learning, and Quan...

Quantum computing employs some quantum phenomena to process information. It has been hailed as the future of computing but it is plagued by serious hurdles when it comes to its practical realization. MemComputing is a new paradigm that instead employs non-quantum dynamical systems and exploits time non-locality (memory) to compute. It can be effici...

The Peltier effect is the reverse phenomenon of the Seebeck effect, and has been observed experimentally in nanoscale junctions. However, despite its promising applications in local cooling of nanoelectronic devices, the role of strong electron correlations on such a phenomenon is still unclear. Here, by analyzing the thermoelectric properties of q...

Custodial symmetries are common in the standard model of particle physics. They arise when quantum corrections to a parameter are proportional to the parameter itself. Here, we show that a custodial symmetry of the chiral type is also present in a classical Su-Schrieffer-Heeger (SSH) electrical circuit with memory. In the absence of memory, the SSH...

Voltage-controlled magnetic anisotropy (VCMA) is a low-energy alternative to manipulate the ferromagnetic state, which has been recently also considered in antiferromagnets (AFMs). Here, we theoretically demonstrate that VCMA can be used to excite linear and parametric resonant modes in easy-axis AFMs with perpendicular anisotropy, thus opening the...

Information processing and storage by the same physical system is emerging as a promising alternative to traditional computing platforms. In turn, this requires the realization of elementary units the memory content of which can be easily tuned and controlled. Here, we introduce a polariton-based quantum memristor where the memristive nature arises...

MemComputing is a new computing paradigm that employs time non-locality (memory) to both process and store information. This book, written by the originator of this paradigm, explains the main ideas behind MemComputing, its theoretical foundations, and shows its applicability to a wide variety of combinatorial optimization problems, Machine Learnin...

Voltage controlled magnetic anisotropy (VCMA) is a low-energy alternative to manipulate the ferromagnetic state, which has been recently considered also in antiferromagnets (AFMs). Here, we theoretically demonstrate that VCMA can be used to excite linear and parametric resonant modes in easy-axis AFMs with perpendicular anisotropy, thus opening the...

Neural networks (NNs) representing quantum states are typically trained using Markov chain Monte Carlo sampling and gradient-based methods. However, such approaches cannot easily extract the global (long-range) structure of quantum states, which is essential for certain problems. Here, we propose to use a mode-assisted training that provides global...

In the era of Big Data and Internet of Things (IoT), information security has emerged as an essential system and application metric. The information exchange among the ubiquitously connected smart electronic devices requires functioning reliably in harsh environments, which highlights the need for securing the hardware root of trust. In this work,...

Spin glasses are notoriously difficult to study both analytically and numerically due to the presence of frustration and metastability. Their highly non-convex landscapes require collective updates to explore efficiently. Currently, most state-of-the-art algorithms rely on stochastic spin clusters to perform non-local updates, but such “cluster alg...

The deep extension of the restricted Boltzmann machine (RBM), known as the deep Boltzmann machine (DBM), is an expressive family of machine learning models which can serve as compact representations of complex probability distributions. However, jointly training DBMs in the unsupervised setting has proven to be a formidable task. A recent technique...

Custodial symmetries are common in the Standard Model of particle physics. They arise when quantum corrections to a parameter are proportional to the parameter itself. Here, we show that a custodial symmetry of the chiral type is also present in a classical Su-Schrieffer-Heeger (SSH) electrical circuit with memory (memcircuit). In the absence of me...

The end of Moore’s law for CMOS technology has prompted the search for low-power computing alternatives, resulting in promising approaches such as nanomagnetic logic. However, nanomagnetic logic is unable to solve a class of interesting problems efficiently, as it only allows for forward computing, due to the need for clocking and/or thermal anneal...

Information processing and storing by the same physical system is emerging as a promising alternative to traditional computing platforms. In turn, this requires the realization of elementary units whose memory content can be easily tuned and controlled. Here, we introduce a polariton-based quantum memristor where the memristive nature arises from t...

Emerging brain-inspired neuromorphic computing paradigms require devices that can emulate the complete functionality of biological synapses upon different neuronal activities in order to process big data flows in an efficient and cognitive manner while being robust against any noisy input. The memristive device has been proposed as a promising cand...

Surface plasmon enhanced processes and hot-carrier dynamics in plasmonic nanostructures are of great fundamental interest to reveal light-matter interactions at the nanoscale. Using plasmonic tunnel junctions as a platform supporting both electrically- and optically excited localized surface plasmons, we report a much greater (over 1000x) plasmonic...

Digital memcomputing machines (DMMs) are a novel, non-Turing class of machines designed to solve combinatorial optimization problems. They can be physically realized with continuous-time, non-quantum dynamical systems with memory (time non-locality), whose ordinary differential equations (ODEs) can be numerically integrated on modern computers. Sol...

It is believed that the $\pm J$ Ising spin-glass does not order at finite temperatures in dimension $d=2$. However, using a graphical representation and a contour argument, we prove rigorously the existence of a finite-temperature phase transition in $d\geq 2$ with $T_c \geq 0.4$. In the graphical representation, the low-temperature phase allows fo...

While resistors with memory, sometimes called memristive elements (such as ReRAM cells), are often studied under conditions of periodic driving, little attention has been paid to the Fourier features of their memory response (hysteresis). Here we demonstrate experimentally that the hysteresis of memristive systems can be unambiguously distinguished...

Surface plasmon enhanced processes and hot-carrier dynamics in plasmonic nanostructures are of great fundamental interest to reveal light-matter interactions at the nanoscale. Using plasmonic tunnel junctions as a platform supporting both electrically and optically excited localized surface plasmons, we report a much greater (over 1000× ) plasmonic...

Novel computational paradigms may provide the blueprint to help solving the time and energy limitations that we face with our modern computers, and provide solutions to complex problems more efficiently (with reduced time, power consumption and/or less device footprint) than is currently possible with standard approaches. Spintronics offers a promi...

A simple and unambiguous test has been recently suggested [J. Phys. D: Applied Physics, 52, 01LT01 (2018)] to check experimentally if a resistor with memory is indeed a memristor, namely a resistor whose resistance depends only on the charge that flows through it, or on the history of the voltage across it. However, although such a test would repre...

The deep extension of the restricted Boltzmann machine (RBM), known as the deep Boltzmann machine (DBM), is an expressive family of machine learning models which can serve as compact representations of complex probability distributions. However, jointly training DBMs in the unsupervised setting has proven to be a formidable task. A recent technique...

Spin glasses are notoriously difficult to study both analytically and numerically due to the presence of frustration and multiple metastable states. Their highly non-convex landscape requires collective updates to explore efficiently. This is currently done through stochastic cluster algorithms, though they lack general efficiency. Here, we introdu...

Digital memcomputing machines (DMMs) are a novel, non-Turing class of machines designed to solve combinatorial optimization problems. They can be physically realized with continuous-time, non-quantum dynamical systems with memory (time non-locality), whose ordinary differential equations (ODEs) can be numerically integrated on modern computers. Sol...

Measuring local temperatures of open systems out of equilibrium is emerging as a novel approach to study the local thermodynamic properties of nanosystems. An operational protocol has been proposed to determine the local temperature by coupling a probe to the system and then minimizing the perturbation to a certain local observable of the probed sy...

The end of Moore's law for CMOS technology has prompted the search for low-power computing alternatives, resulting in several promising proposals based on magnetic logic[1-8]. One approach aims at tailoring arrays of nanomagnetic islands in which the magnetostatic interactions constrain the equilibrium orientation of the magnetization to embed logi...

Boolean satisfiability is a propositional logic problem of interest in multiple fields, e.g., physics, mathematics, and computer science. Beyond a field of research, instances of the SAT problem, as it is known, require efficient solution methods in a variety of applications. It is the decision problem of determining whether a Boolean formula has a...

Boolean satisfiability is a propositional logic problem of interest in multiple fields, e.g., physics, mathematics, and computer science. Beyond a field of research, instances of the SAT problem, as it is known, require efficient solution methods in a variety of applications. It is the decision problem of determining whether a Boolean formula has a...

While resistors with memory, sometimes called memristive elements (such as ReRAM cells), are often studied under conditions of periodic driving, little attention has been paid to the Fourier features of their memory response (hysteresis). Here we demonstrate experimentally that the hysteresis of memristive systems can be unambiguously distinguished...

Measuring local temperatures of open systems out of equilibrium is emerging as a novel approach to study the local thermodynamic properties of nanosystems. An operational protocol has been proposed to determine the local temperature by coupling a probe to the system and then minimizing the perturbation to a certain local observable of the probed sy...

It has been suggested that all resistive‐switching memory cells are memristors. The latter are hypothetical, ideal devices whose resistance, as originally formulated, depends only on the net charge that traverses them. Recently, an unambiguous test has been proposed to determine whether a given physical system is indeed a memristor or not. Here, su...

Restricted Boltzmann machines (RBMs) are a powerful class of generative models, but their training requires computing a gradient that, unlike supervised backpropagation on typical loss functions, is notoriously difficult even to approximate. Here, we show that properly combining standard gradient updates with an off-gradient direction, constructed...

As demands for computing have been continuously increasing, various solutions have been proposed. Some approaches deal with improving algorithms and software programs, mainly through the tuning of advanced heuristics and learning methods. Some other tackle the problem by providing specialized hardware to enhance computation. Other revolutionary app...

Restricted Boltzmann machines (RBMs) are a powerful class of generative models, but their training requires computing a gradient that, unlike supervised backpropagation on typical loss functions, is notoriously difficult even to approximate. Here, we show that properly combining standard gradient updates with an off-gradient direction, constructed...

Recent work on quantum annealing has emphasized the role of collective behavior in solving optimization problems. By enabling transitions of clusters of variables, such solvers are able to navigate their state space and locate solutions more efficiently despite having only local connections between elements. However, collective behavior is not excl...

Novel computational paradigms may provide the blueprint to help solving the time and energy limitations that we face with our modern computers, and provide solutions to complex problems more efficiently (with reduced time, power consumption and/or less device footprint) than is currently possible with standard approaches. Spintronics offers a promi...

The temperature of a physical system is operationally defined in physics as "that quantity which is measured by a thermometer" weakly coupled to, and at equilibrium with the system. This definition is unique only at global equilibrium in view of the zeroth law of thermodynamics: when the system and the thermometer have reached equilibrium, the "the...

The temperature of a physical system is operationally defined in physics as “that quantity which is measured by a thermometer” weakly coupled to, and at equilibrium with the system. This definition is unique only at global equilibrium in view of the zeroth law of thermodynamics: when the system and the thermometer have reached equilibrium, the “the...

Wang et al. claim [J. Appl. Phys. 125, 054504 (2019)] that a current-carrying wire interacting with a magnetic core represents a memristor. Here, we demonstrate that this claim is false. We first show that such memristor "discovery" is based on incorrect physics, which does not even capture basic properties of magnetic core materials, such as their...

It has been suggested that all resistive-switching memories are memristors. The latter are hypothetical, ideal devices whose resistance, as originally formulated, depends only on the net charge that traverses them. Recently, an unambiguous test has been proposed [J. Phys. D: Appl. Phys. 52, 01LT01 (2019)] to determine whether a given physical syste...

Memcomputing is a novel computing paradigm that employs time non-locality (memory) to solve combinatorial optimization problems. It can be realized in practice by means of non-linear dynamical systems whose point attractors represent the solutions of the original problem. It has been previously shown that during the solution search digital memcompu...

An analysis of the literature shows that there are two types of non-memristive models that have been widely used in the modeling of so-called "memristive" neural networks. Here, we demonstrate that such models have nothing in common with the concept of memristive elements: they describe either non-linear resistors or certain bi-state systems, which...

We report the confirmed occurrence of Fowler-Nordheim hole tunneling in p-4H-SiC metal-oxide-semiconductor capacitor structures. The effective mass for holes in the oxide is found to be in the range of 0.35m-0.52m, where m is the free electron mass. A fundamental process in the description of the current-voltage (I-V) characteristic of a metal-diel...

Chaos is an ubiquitous and fundamental phenomenon with a wide range of features pointing to a similar phenomenology. Although apparently distinct, it is natural to ask if all these features emerge from a unifying principle. Recently, it was realized that all continuous-time stochastic dynamical systems (DSs) — the most relevant in physics because n...

Memcomputing is a novel computing paradigm that employs time non-local dynamical systems to compute with and in memory. The digital version of these machines [digital memcomputing machines or (DMMs)] is scalable, and is particularly suited to solve combinatorial optimization problems. One of its possible realizations is by means of standard electro...

Digital memcomputing machines (DMMs) are a class of computational machines designed to solve combinatorial optimization problems. A practical realization of DMMs can be accomplished via electrical circuits of highly non-linear, point-dissipative dynamical systems engineered so that periodic orbits and chaos can be avoided. A given logic problem is...