Bulletin of the American Physical Society
APS March Meeting 2023
Volume 68, Number 3
Las Vegas, Nevada (March 5-10)
Virtual (March 20-22); Time Zone: Pacific Time
Session T02: Information in ThermodynamicsFocus
|
Hide Abstracts |
Sponsoring Units: GSNP Chair: Marc Serra Garcia, ETH Zurich Room: Room 125 |
Thursday, March 9, 2023 11:30AM - 12:06PM |
T02.00001: Materializing Cognition: Information processing with cognitive matter Invited Speaker: Wilfred van de Wiel Throughout history, man has exploited matter to carry out tasks well beyond his biological constraints. Starting from primitive tools with functionality solely derived from shape and structure, we have moved on to responsive matter that can change its properties upon external stimulus and even further to adaptive matter that can change its response depending on the environment. One of the grand scientific and intellectual challenges is to make matter that can actually learn. Such matter’s behavior would not only depend on the here and now, but also on its past. It would have memory, autonomously interact with its environment and self-regulate its action. We may call such matter ‘intelligent’. |
Thursday, March 9, 2023 12:06PM - 12:18PM |
T02.00002: Development of Property-Performance Links for Vanadium Dioxide Nonlinear Dynamical Memristors via Local Activity Modeling Timothy D Brown, Stephanie M Bohaichuk, Mahnaz Islam, Suhas Kumar, R. Stanley Williams The absence of properties-performance relations is a crucial limitation for translating the surging interest in nonlinear electronic devices into their large-scale commercial deployment in neuromorphic networks. In order to develop systematic reverse- and forward-design principles for neuromorphic materials, materials and device engineers need a unified framework, within which “exotic” dynamical behaviors (negative differential resistance, persistent spiking oscillations, spontaneous localization of current channels) are linked with nonlinear transport physics. In this work, we advance the dynamical theory of Local Activity as such a unified framework for modeling neuromorphic behavior and extracting nonlinear materials properties in VO2/SiN electro-thermal memristors. Using a recent physical recontextualization of Local Activity in terms of competition between Joule Heating and Newtonian Cooling thermal feedback, we develop systematic least-squares fitting procedures to experimental steady-state and dynamical data to independently characterize the temperature-dependent electrical and thermal transport material properties of the VO2 memristors, and directly link the data to the underlying electro-thermal nonlinearities. Beyond accurately reproducing the experimental data, the modeling procedure is (1) predictive, in terms of producing surprising but independently-verifiable electro-thermal material characterization, and (2) informative, in terms of linking features of the steady state curve to both dynamical behavior and to the underlying nonlinear electro-thermal transport. This recontextualization of Local Activity compact modeling in terms of systematic fitting procedures has been a missing link in predictively connecting nonlinear electronic material properties to neuromorphic device performance, and should enable rapid screening of candidate materials manifesting nonlinear electrical and thermal transport physics. |
Thursday, March 9, 2023 12:18PM - 12:30PM |
T02.00003: Second Law for Active Heat Engines Arya Datta, Patrick Pietzonka, Andre Barato Macroscopic cyclic heat engines have been a major motivation for the emergence of thermodynamics. In the past decade, cyclic heat engines that have large fluctuations and operate at finite time were studied within the more modern framework of stochastic thermodynamics. The second law for such heat engines states that the efficiency cannot be larger than the Carnot efficiency. The concept of cyclic active heat engines for a system in the presence of hidden dissipative degrees of freedom, also known as a nonequilibrium or active reservoir, has also been studied in theory and experiment. Such active engines show rather interesting behavior such as an "efficiency" larger than the Carnot bound. They are also likely to play an important role in future developments, given the ubiquitous presence of active media. However, a general second law for cyclic active heat engines has been lacking so far. Here, by using a known inequality in stochastic thermodynamics for the excess entropy, we obtain a general second law for active heat engines, which does not involve the energy dissipation of the hidden degrees of freedom and is expressed in terms of quantities that can be measured directly from the observable degrees of freedom. Besides heat and work, our second law contains an information-theoretic term, which allows an active heat engine to extract work beyond the limits valid for a passive heat engine. To obtain a second law expressed in terms of observable variables in the presence of hidden degrees of freedom, we introduce a coarse-grained excess entropy and prove a fluctuation theorem for this quantity. |
Thursday, March 9, 2023 12:30PM - 12:42PM |
T02.00004: Information thermodynamics of the transition-path ensemble David A Sivak, Miranda D Louwerse The reaction coordinate describing a transition between reactant and product is a fundamental concept in the theory of chemical reactions. Within transition-path theory, a quantitative definition of the reaction coordinate is found in the committor, which is the probability that a trajectory initiated from a given microstate first reaches the product before the reactant. Here we develop an information-theoretic origin for the committor and show how selecting transition paths from a long ergodic equilibrium trajectory induces entropy production which exactly equals the information that system dynamics provide about the reactivity of trajectories. This equality of entropy production and dynamical information generation also holds at the level of arbitrary individual coordinates, providing parallel measures of the coordinate's relevance to the reaction, each of which is maximized by the committor. |
Thursday, March 9, 2023 12:42PM - 12:54PM |
T02.00005: Empirical Modeling of Superparamagnetic Magnetic Tunnel Junctions with Application to Probabilistic Computing Liam A Pocher, Sidra Gibeault, Advait Madhavan, Mark D Stiles, Matthew Daniels, Nhat-Tan Phan, Philippe Talatchian, Ursula Ebels, Daniel P Lathrop With the end of Moore's Law as we approach atomic scales, new and innovative ways are required to perform computation. Novel paradigms including beyond von Neumann architectures are one approach. An example is probabilistic computing, which leverages internal stochastic behavior for computation, one application being simulated annealing. To design next generation hardware architectures, we require high fidelity models capturing internal physics. In this work we describe an empirical model based on the Langevin equation that accurately captures quantitative metrics associated with one probabilistic bit realized in a superparamagnetic tunnel junction. We show how our model can be reduced to a one degree of freedom massless model capturing dynamics with high fidelity when compared to experimental data from a superparamagnetic tunnel junction. We then show how this one degree of freedom model can be used in computer design software enabling rapid prototyping of next generation computer architectures. |
Thursday, March 9, 2023 12:54PM - 1:06PM |
T02.00006: Temperature sensitivity of true random number generation in stochastic magnetic actuated random transducer devices Laura Rehm, Corrado Capriata, Shashank Misra, J. Darby Smith, Mustafa Pinarbasi, Bengt Gunnar Malm, Andrew D Kent True random number sources are of great interest for numerous applications such as cryptography and Monte Carlo simulations. Here we show that perpendicularly magnetized magnetic tunnel junctions operated in the ballistic switching limit (ns duration pulses) have great potential for such applications due to the stochasticity of their spin-transfer-torque driven reversal. In the ballistic limit the resulting junction state is random mainly because of the thermal distribution of the initial magnetization state. We denote this a stochastic magnetic actuated random transducer (SMART) device because a pulse activates the junction to generate a random bit, much like a coin flip. We find that the experimentally obtained switching probability characteristics can be well described by a simple macrospin model. We also show that our SMART devices with a thermal stability factor (Δ = 39) operated in the ballistic limit are much less sensitive to temperature variations than the same device operated with longer pulses. In addition, we investigate the stochastic nature of our SMART devices by comparing their statistics to Bernoulli trials and show that we can successfully sample a uniform distribution1, which is commonly used in computations that require random numbers. Our results demonstrate that SMART devices are a great candidate for true random number generation due to their easily controllable characteristics, while being relatively robust towards environmental changes. |
Thursday, March 9, 2023 1:06PM - 1:18PM |
T02.00007: Probabilistic computing and stochastic devices Shashank Misra, Christopher R Allemang, Laura Rehm, Andrew D Kent, Jean Anne C Incorvia, Leslie C Bland, Catherine Schuman, Suma G Cardwell, J. Darby Smith, J. Bradley Aimone Probabilistic computing considers a range of applications, all which share a requirement for a high volume of samples from different probability distributions. For example, modeling some nuclear physics using Monte Carlo codes results in half their run time spent generating uniform pseudo random numbers, and significant computational overhead transforming those numbers to sample relevant distributions. Thus, there is a benefit in connecting the stochasticity of a device to statistical sampling in a way that makes sampling ubiquitous and cheap, in time and energy, which may further motivate the development of algorithms that continue to shift the burden from calculation to sampling. This talk focuses on statistical analysis of device bitstreams based on their ability to generate quality statistical samples. We focus on understanding the implications of these requirements on two promising devices – magnetic tunnel junctions and tunnel diodes. We conclude with resource estimates for circuits capable of efficiently producing samples for large probabilistic calculations. |
Thursday, March 9, 2023 1:18PM - 1:30PM |
T02.00008: Small coupling strength is not the necessary condition for strong-coupling effects to disappear Jong-Min Park, Hyunggyu Park, Jae Sung Lee Lots of dynamics of microscopic systems are described by the Langevin equation that contains no terms depending on the interaction between the system and the environment, although the system energy scale is comparable to that of the interaction. One might guess that this weak-coupling feature originates from a small coupling strength. But the small coupling strength limit is unreasonable because it leads the system to evolve deterministically with decoupled from the environment. In this presentation, we show the condition for strong-coupling effects to disappear regardless of the coupling strength. For this purpose, we consider a strongly-coupled system evolving deterministically coupled to the bath particles in the Langevin thermostat. By taking the limit as the relaxation time of the bath goes to zero, we derive the stochastic equation of motion for the reduced dynamics. This equation contains two interaction-dependent terms, (i) an additional potential and (ii) an effective damping tensor. By revealing the condition where the conventional weakly-coupled Langevin equation is restored even with a finite coupling strength, we confirm that a small coupling is not the necessary condition for strong-coupling effects to vanish. We verify our result numerically in systems with various forms of the interaction potential. We show that when the condition is satisfied, the interaction form has no influence on the equation of motion for the reduced system. |
Thursday, March 9, 2023 1:30PM - 1:42PM |
T02.00009: Quantum Wheatstone bridge Kasper Poulsen The quantum mechanical effects of entanglement and interference are the drivers of modern quantum technologies. Combining these with the unidirectional dynamics of cold thermal baths yield additional possibilities in areas such as error correction, heat control, and metrology. One example of such a device is a quantum version of the classical Wheatstone bridge. The quantum Wheatstone bridge is comprised of four spins coupled to thermal baths at the boundaries and exploits quantum effects to gain an enhanced sensitivity to an unknown coupling strength. The four spins are arranged in a double slit formation allowing an entangled bell state to develop on the interface. A controllable coupling is varied until the bell state is destroyed due to destructive interference which is called the balance point. At the balance point the unknown coupling is the same as the controllable coupling which is robust towards calibration errors and decoherence. Alternatively, the balance point can be found by maximizing the spin transport between the two baths. This makes for a device for measuring an unknown coupling working through quantum effects and driven by the non-unitarity of the thermal baths. |
Thursday, March 9, 2023 1:42PM - 1:54PM |
T02.00010: Unitary k-designs from random number-conserving quantum circuits Sumner Hearth, Michael Flynn, Anushya Chandran, Christopher R Laumann In a number conserving quantum system the kth-Renyi entropy can be extracted from statistical correlations between measurements in random bases consistent with the conservation law. Unlike systems without a conserved number, the entropy cannot be extracted from the measurements of circuits composed of spatially non-overlapping unitaries. Instead, the minimal circuit has a brick layer structure with depth scaling with the square of the linear dimension for a maximal error in the purity of order one. This scaling of the depth is a direct a consequence of the diffusive dynamics of the conserved number. Surprisingly, we observe this scaling even when the number density in the state is uniform, e.g. in states with hyperuniform particle distributions and uniformly delocalized particles. |
Thursday, March 9, 2023 1:54PM - 2:06PM |
T02.00011: Quasi-Skin Modes in a Nonlinear non-Hermitian Lattice Hamed Ghaemidizicheh, Cem Yuce, Hamidreza Ramezani An extensive number of eigenstates become localized at the edge of a one-dimensional linear non-Hermitian lattice due to the non-Hermitian skin effect (NHSE). A quasi-skin mode appears in the semi-infinite lattice as a result of NHSE and can almost preserve its profile up to a survival time in the finite lattices. A nonlinear extension of NHSE has recently attracted some attention. In our talk, we obtain quasi-skin modes under the semi-infinite boundary conditions and explore their survival times in the finite nonlinear lattice with open edges. We show that the survival time increases with lattice size but decreases dramatically with the disorder. |
Thursday, March 9, 2023 2:06PM - 2:18PM |
T02.00012: Long-Lived Solitons, anomalous dynamics and equilibration in the classical Heisenberg chain Thomas Bilitewski, Adam J McRoberts, Roderich Moessner, Masudul Haque The search for departures from standard hydrodynamics in many-body systems has yielded a number of promising leads, especially in low dimensions. Here, we study one of the simplest classical interacting lattice models, the nearest-neighbor Heisenberg chain, with temperature as the tuning parameter. Our numerics expose strikingly different spin dynamics between the antiferromagnet, where it is largely diffusive, and the ferromagnet, where we observe strong evidence either of spin superdiffusion or an extremely slow crossover to diffusion. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700