Search

General quadratically constrained quadratic programs (QCQPs) are challenging to solve as they are known to be NP-hard. A popular approach to approximating QCQP solutions is to use semidefinite programming (SDP) relaxations. It is well-known that the optimal value $\eta$ of the SDP relaxation problem bounds the optimal value $\zeta$ of the QCQP from below, i.e., $\eta \leq \zeta$. The two problems are considered equivalent if $\eta = \zeta$. In the recent paper by Arima, Kim and Kojima [arXiv:2409.07213], a class of QCQPs that are equivalent to their SDP relaxations are proposed with no condition imposed on the quadratic objective function, which can be chosen arbitrarily. In this work, we explore the construction of various QCQP instances within this class to complement the results in [arXiv:2409.07213]. Specifically, we first construct QCQP instances with two variables and then extend them to higher dimensions. We also discuss how to compute an optimal QCQP solution from the SDP relaxation., Comment: 23 pages, 13 figures

Antiferromagnets with spin splitting in the electronic structure are desired for novel spintronic phenomena in materials with nearly zero net magnetization. One of the simplest spin-split antiferromagnetic states, the $p$-wave magnet, is proposed as a result of collective instability of electron gases. Following a more recent theoretical proposal without strong electron correlations, we here report a material with the symmetry constraints for $p$-wave magnetism of conduction electrons in momentum space, based on a texture of magnetic moments in direct space. Our resonant X-ray scattering reveals coplanar, lattice-locked antiferromagnetism that satisfies the required conditions for a $p$-wave magnet: it breaks space inversion, but preserves time-reversal ($T$) symmetry up to a positional shift in direct space. Consistent with theoretical predictions, the electric conductivity is characteristically anisotropic in this $p$-wave magnet. In conducting magnets, the coupling of magnetic spins and freely moving electrons favors small distortions of the $p$-wave state, slightly breaking the $T$ symmetry and inducing a tiny net magnetization. In our material, this gentle symmetry breaking induces an anomalous Hall effect (AHE) with a giant anomalous Hall conductivity for a bulk antiferromagnet, $\sigma_{xy}>600\,$S/cm (Hall angle $>3\,\%$). The $p$-wave magnet has characteristic spin-nodal planes, and such a giant AHE can be attributed to hybridization of electron bands around these nodal regions due to $T$ breaking., Comment: 10 pages, 4 figures

Double perovskites containing heavy transition metal ions are an important family of compounds for the study of the interplay between electron correlation and spin-orbit coupling. Here, by combining magnetic susceptibility, heat capacity, and neutron scattering measurements, we investigate the dipolar and quadrupolar correlations in two prototype rhenium-based double perovskite compounds, Ba$_2$YReO$_6$ and Ba$_2$ScReO$_6$. A type-I dipolar antiferromagnetic ground state with a propagation vector $\mathbf{q} = (0, 0, 1)$ is observed in both compounds. At temperatures above the magnetic transitions, a quadrupolar ordered phase is identified. Weak spin excitations, which are gapped at low temperatures and softened in the correlated paramagnetic phase, are explained using a minimal model that considers both the dipolar and quadrupolar interactions. At larger wavevectors, we observe dominant phonon excitations that are well described by density functional calculations., Comment: 13 pages, 11 figures

Quantized vortices in type-II superconductors provide insights into the mechanisms of superconductivity. However, the generation of antivortices, characterized by magnetization antiparallel to the external magnetic field, remains less understood. In this study, we investigate Sn-Pb solder, a superconductor with phase-separated Sn and Pb phases, and report the observation of both vortices and antivortices. Scanning SQUID (superconducting quantum interference device) microscopy revealed the presence of both vortices and antivortices, while magneto-optical imaging demonstrated flux avalanches. Our results demonstrate that Sn in Sn-Pb solder behaves as a type-II superconductor when magnetic fluxes are trapped, despite bulk Sn being a type-I superconductor with a transition temperature (TcSn) of 3.7 K. Our findings suggest that the size effect and proximity effect with Pb contribute synergistically to induce type-II superconductivity in Sn Notably, vortices were observed at temperatures as high as 5 K, exceeding the bulk TcSn. Furthermore, the interplay between the type-I superconducting Pb phase and the type-II superconducting Sn phase results in the generation of antivortices, providing a mechanism to accommodate excess magnetic flux. This study shed light on new research on composites combining type-I and type-II superconductors., Comment: 27 pages, 15 figures

Nodal-line semimetals are a class of topological materials hosting one dimensional lines of band degeneracy. Kramers nodal-line (KNL) metals/semimetals have recently been theoretically recognized as a class of topological states inherent to all non-centrosymmetric achiral crystal lattices. We investigate the electronic structure of candidate KNL semimetal YAuGe by angle-resolved photoemission spectroscopy (ARPES) and quantum oscillations as well as by density functional theory (DFT) calculations. DFT has revealed that YAuGe hosts KNLs on the G-A-L-M plane of the Brillouin zone, that are protected by the time reversal and mirror-inversion symmetries. Through ARPES and quantum oscillations we identify signatures of hole bands enclosing the G point, and the observed splitting of quantum oscillation frequency with angle is attributed to spin-orbit-coupling-induced band splitting away from the KNLs. Furthermore, we show that the degeneracy of the nodal lines along the G-A line is lifted by the time-reversal-symmetry breaking when the Y is substituted by magnetic R ions (R = rare earth). This becomes a source of Berry curvature and contributes to the anomalous Hall effect in magnetic RAuGe. These findings establish RAuGe as a new class of KNL semimetals offering significant potential for engineering of anomalous magnetotransport properties via magnetic rare-earth substitution., Comment: Main: 8 pages; 4 figures, Supplementary: 7 pages; 9 figures; 3 tables

Recent studies on plant disease diagnosis using machine learning (ML) have highlighted concerns about the overestimated diagnostic performance due to inappropriate data partitioning, where training and test datasets are derived from the same source (domain). Plant disease diagnosis presents a challenging classification task, characterized by its fine-grained nature, vague symptoms, and the extensive variability of image features within each domain. In this study, we propose the concept of Discriminative Difficulty Distance (DDD), a novel metric designed to quantify the domain gap between training and test datasets while assessing the classification difficulty of test data. DDD provides a valuable tool for identifying insufficient diversity in training data, thus supporting the development of more diverse and robust datasets. We investigated multiple image encoders trained on different datasets and examined whether the distances between datasets, measured using low-dimensional representations generated by the encoders, are suitable as a DDD metric. The study utilized 244,063 plant disease images spanning four crops and 34 disease classes collected from 27 domains. As a result, we demonstrated that even if the test images are from different crops or diseases than those used to train the encoder, incorporating them allows the construction of a distance measure for a dataset that strongly correlates with the difficulty of diagnosis indicated by the disease classifier developed independently. Compared to the base encoder, pre-trained only on ImageNet21K, the correlation higher by 0.106 to 0.485, reaching a maximum of 0.909., Comment: 8 pages, 2 figures, 3 tables. Accepted at 4th Annual AAAI Workshop on AI to Accelerate Science and Engineering (AI2ASE)

We present a brief history of how the famous Maximum Entropy Principle was used as closure of moments of the Boltzmann equation. In particular, we want to remark on the important role of two fundamental papers by Wolfgang Dreyer, one in the classical framework and one in a relativistic context, to use this principle and to compare the result with the macroscopic theory of Rational Extended Thermodynamics., Comment: 28 pages, 0 figures, Paper accepted to a Volume dedicated to the memory of Wolfgang Dreyer (Springer 2025)

We report the magnetization and the $^{51}$V NMR measurements in the polar magnet GaV$_4$Se$_8$ in which a magnetic skyrmion lattice appears in the structural domain with the polar axis parallel to the magnetic field. Although we successfully separate the $^{51}$V NMR signals in the domain from those in the other structural domains, only the high-frequency region of NMR spectrum is observed due to a significant increase of the spin-echo decay rate in the low-frequency region. In $B_{\rm{ext}}\parallel[111]$, we find the NMR signals from the supermagnetized cycloidal state in the parallel domains as well as from the conical state in domains where the polar axis is tilted from the magnetic field. No NMR signal from the skyrmion lattice state is observed, suggesting a significant increase of the decay rate by additional low-energy excitations caused by dynamics of the skyrmions. In $B_{\rm{ext}}\parallel[001]$, where all the structural domains are magnetically equivalent, multiple NMR peaks converge into one peak at the saturation magnetic field. This field dependence is explained by the closing of magnetic cones as approaching the forced-ferromagnetic state., Comment: 37 pages, 20 figures

Although the average life expectancy in Japan has been increasing in recent years, the problem of the large gap between healthy life expectancy and average life expectancy is still unresolved. Among the factors that lead to the need for nursing care, injuries due to falls account for a certain percentage of the total. In this paper, we developed boots that can provide external disturbance to the ankle with pneumatic gel muscles (PGM) while walking. We experimented using an angular velocity and acceleration of the heel as evaluation indices to evaluate the effectiveness of fall prevention training using this device, which is smaller and more wearable than conventional devices. In this study, we confirmed that the developed system has enough training intensity to significantly affect the gait waveform., Comment: Part of proceedings of 6th International Conference AsiaHaptics 2024

A new expression is here proposed for the production term Q in the Boltzmann Equation. This new expression satisfies the H-Theorem (non approximated) and, in the transition to Ordinary Thermodynamics gives a result which is independent on the number N of the moments which are used both if the Maxwellian Iteration or the Eckart method are used.

Electrons in localized 4f orbitals around lanthanide elements in solids are dominated by strong spin-orbit coupling. The resultant J multiplets are further split by the crystalline electric field, where J denotes the total angular momentum quantum number, and hence exhibit anisotropic spatial distribution. The anisotropy of spin-orbit entangled 4f electrons is closely linked to their physical properties. Here, we demonstrate the visualization of 4f electrons in pyrochlore oxides A2Ir2O7 (A = Pr, Nd, and Eu) by combining high-photon-energy X-ray diffraction and valence electron density (VED) analysis based on the core differential Fourier synthesis (CDFS) method. The observed VED distributions around the A site roughly agree with point-charge calculations, considering the relativistic spin-orbit coupling and the trigonal crystal electric field under the LS coupling scheme. The CDFS-based VED observation method can be applied to the analysis of the 4f state of various crystalline materials., Comment: 19 pages, 3 figures, SI

We respond to the "Comment on "Reconsidering the nonlinear emergent inductance: time-varying Joule heating and its impact on the AC electrical response"", submitted by Yokouchi et al. (arXiv:2407.15682), which refers to our preprint by Furuta et al (arXiv:2407.00309). In our preprint, we argued that the nonlinear impedance reported for materials with noncollinear magnetic textures, Gd$_3$Ru$_4$Al$_{12}$, YMn$_6$Sn$_6$, and FeSn$_2$, can be attributed to the considerable contribution of Joule-heating-induced AC electrical response, rather than the emergent electric field (EEF) due to the current-induced-dynamics of magnetic textures. In the comment by Yokouchi et al., the authors presented new data and concluded that the nonlinear impedance previously reported for Gd$_3$Ru$_4$Al$_{12}$ and YMn$_6$Sn$_6$ was not due to Joule heating but to the EEF. After reviewing their data and arguments, we found that (i) they misunderstood the applicability of the Joule heating model discussed in our preprint (arXiv:2407.00309), (ii) the data they presented in the comment are affected by extrinsic superconductivity caused during the microfabrication of specimens using focused ion beam (FIB) methods, and (iii) they misinterpreted their new data regarding the magnetic-field-angle dependence of the nonlinear impedance. Thus, we maintain that the nonlinear impedance reported in the literature and the comment includes a considerable impact from Joule heating. Furthermore, we demonstrate that for Gd$_3$Ru$_4$Al$_{12}$, the cross-sectional area dependence of the nonlinear impedance refutes the EEF scenario, and that the sign change of nonlinear resistance observed by Yokouchi \textit{et al.}~is not allowed from the perspective of stability associated with the current-driven EEF., Comment: 34 pages, 12 figures; reply to arXiv:2407.15682

The Hall effect is one of the most fundamental but elusive phenomena in condensed matter physics due to the rich variety of underlying mechanisms. Here we report an exceptionally large Hall effect in a frustrated magnet GdCu$_2$ with high conductivity. The Hall conductivity at the base temperature is as high as 4 x 10$^4$ $\Omega^{-1}$cm$^{-1}$ and shows abrupt sign changes under magnetic fields. Remarkably, the giant Hall effect is rapidly suppressed as the longitudinal conductivity is lowered upon increasing temperature or introducing tiny amount of quenched disorder. Our systematic transport measurements together with neutron scattering measurements and ab initio band calculations indicate that the unusual Hall effect can be understood in terms of spin-splitting induced emergence/disappearance of Fermi pockets as well as skew scattering from spin-chiral cluster fluctuations in a field-polarized state. The present study demonstrates complex interplay among magnetization, spin-dependent electronic structure, and spin fluctuations in producing the giant Hall effect in highly conductive frustrated magnets.

We investigate exact semidefinite programming (SDP) relaxations for the problem of minimizing a nonconvex quadratic objective function over a feasible region defined by both finitely and infinitely many nonconvex quadratic inequality constraints (semi-infinite QCQPs). Specifically, we present two sufficient conditions on the feasible region under which the QCQP, with any quadratic objective function over the feasible region, is equivalent to its SDP relaxation. The first condition is an extension of a result recently proposed by the authors (arXiv:2308.05922, to appear in SIAM J. Optim.) from finitely constrained quadratic programs to semi-infinite QCQPs. The newly introduced second condition offers a clear geometric characterization of the feasible region for a broad class of QCQPs that are equivalent to their SDP relaxations. Several illustrative examples, including quadratic programs with ball-, parabola-, and hyperbola-based constraints, are also provided., Comment: 28paages, 6 figures

Magneto-thermal switching (MTS) is a key technology for efficient thermal management. Recently, large MTS with nonvolatility has been observed in Sn-Pb solders [H. Arima et al., Commun. Mater. 5, 34 (2024)] where phase separation, different superconducting transition temperature (Tc) of Sn and Pb, and magnetic-flux trapping are the causes of the nonvolatile MTS. To further understand the mechanism and to obtain the strategy for enhancing switching ratio, exploration of new phase-separated superconductors with nonvolatile MTS is needed. Here, we show that the In52-Sn48 commercial solder is a phase-separated superconducting composite with two Tc and traps vortices after field cooling. A clear signature of nonvolatile MTS was observed at T = 2.5 K. From specific heat analyses, we conclude that the vortices are mainly trapped in the lower-Tc phase ({\gamma}-phase) after field cooling, which is the evidence that vortex trapping also works on achieving nonvolatile MTS in phase-separated superconducting composites., Comment: 14 pages, 5 figures

We show that the convergence rate of the cusp winding spectrum to the Hausdorff dimension of the limit set of a generalized Schottky group with one parabolic generator is polynomial. Our main theorem provides the new phenomenon in which differences in the Hausdorff dimension of the limit set generated by a Markov system cause essentially different results on multifractal analysis. This paper also provides a new characterization of the geodesic flow on the Poinca\'re disc model of two-dimensional hyperbolic space and the limit set of a generalized Schottky group. To prove our main theorem we use thermodynamic formalism on a countable Markov shift, gamma function, and zeta function., Comment: 15 pages

Honeycomb-lattice antiferromagnets have attracted wide attention for exploration of exotic heat transport and their interplay with magnetic excitations. In this work, we have revealed a contrasting behavior in the magneto-thermal conductivity (MTC) between two Co-based honeycomb-lattice magnets Co4M2O9 (M = Nb, Ta), despite their identical lattice structures and quite similar magnetism. Co4Ta2O9 exhibits enhanced MTC of about 550% at 9 T of an in-plane magnetic field, comparable to other honeycomb-magnets, while MTC for Co4Nb2O9 reaches only ~30%. This marked difference is ascribed to distinct features in the field-induced evolution of magnetic excitations that resonantly scatter phonons. This finding sheds light on implicit impacts of non-magnetic ions on thermal transport, and hints at the potential for broad heat-transport tunability while preserving magnetism and lattice structures., Comment: 7 pages, 4 figures, doi.org/10.1103/PhysRevB.110.L041116

A nonlinearly enhanced electrical reactance, $\Im Z$, under a large AC current has been measured to explore emergent inductors, which constitute a new class of inductors based on the spin-transfer torque effect. A nonlinear $\Im Z$ has been observed in conducting magnets that contain noncollinear spin textures and interpreted as the realization of an inductance due to current-induced spin dynamics. However, curious behavior has concomitantly been observed. For instance, the nonlinear $\Im Z$ always has a cutoff frequency of $10^0$--$10^4$ Hz, which is much lower than the resonance frequency of a ferromagnetic domain wall, $\sim$10$^7$ Hz; furthermore, the temperature and magnetic field variations in $\Im Z$ appear to be considerably correlated with those in the temperature derivative of resistance. This behavior appears to be difficult to understand in terms of the current-induced spin dynamics, and therefore, the earlier interpretation of the nonlinear $\Im Z$ should be further verified. Here, we theoretically and experimentally show that time-varying Joule heating and its impact on the AC electrical response can naturally explain these observations. In the experimental approach, we study the nonlinear AC electrical response of two conducting materials that exhibit no magnetic order, CuIr$_2$S$_4$ and 1$T$'-MoTe$_2$. Under time-varying Joule heating, a nonlinearly enhanced $\Im Z$ with the curious behavior mentioned above is observed in both systems. Our study implies that the nonlinear $\Im Z$ previously observed in noncollinear magnets includes a considerable contribution of the Joule-heating-induced apparent AC impedance., Comment: 34 pages, 16 figures

We studied the optical variability of 241 BL Lacs and 83 flat-spectrum radio quasars (FSRQ) from the 4LAC catalog using data from the Tomo-e Gozen Northern Sky Transient Survey, with $\sim$ 50 epochs per blazar on average. We excluded blazars whose optical variability may be underestimated due to the influence of their host galaxy, based on their optical luminosity ($L_O$). FSRQs with $\gamma$-ray photon index greater than 2.6 exhibit very low optical variability, and their distribution of standard deviation of repeated photometry is significantly different from that of the other FSRQs (KS test P value equal to $5 \times 10^{-6}$ ). Among a sample of blazars at any particular cosmological epoch, those with lower $\gamma$-ray luminosity ($L_\gamma$) tend to have lower optical variability, and those FSRQs with $\gamma$-ray photon index greater than 2.6 tend to have low $L_\gamma$. We also measured the structure function of optical variability and found that the amplitude of the structure function for FSRQs is higher than previously measured and higher than that of BL Lacs at multiple time lags. Additionally, the amplitude of the structure function of FSRQs with high $\gamma$-ray photon index is significantly lower than that of FSRQs with low $\gamma$-ray photon index. The structure function of FSRQs of high $\gamma$-ray photon index shows a characteristic timescale of more than 10 days, which may be the variability timescale of the accretion disk. In summary, we infer that the optical component of FSRQs with high $\gamma$-ray photon index may be dominated by the accretion disk., Comment: 22 pages, 13 figures

GPU-based heterogeneous architectures are now commonly used in HPC clusters. Due to their architectural simplicity specialized for data-level parallelism, GPUs can offer much higher computational throughput and memory bandwidth than CPUs in the same generation do. However, as the available resources in GPUs have increased exponentially over the past decades, it has become increasingly difficult for a single program to fully utilize them. As a consequence, the industry has started supporting several resource partitioning features in order to improve the resource utilization by co-scheduling multiple programs on the same GPU die at the same time. Driven by the technological trend, this paper focuses on hierarchical resource partitioning on modern GPUs, and as an example, we utilize a combination of two different features available on recent NVIDIA GPUs in a hierarchical manner: MPS (Multi-Process Service), a finer-grained logical partitioning; and MIG (Multi-Instance GPU), a coarse-grained physical partitioning. We propose a method for comprehensively co-optimizing the setup of hierarchical partitioning and the selection of co-scheduling groups from a given set of jobs, based on reinforcement learning using their profiles. Our thorough experimental results demonstrate that our approach can successfully set up job concurrency, partitioning, and co-scheduling group selections simultaneously. This results in a maximum throughput improvement by a factor of 1.87 compared to the time-sharing scheduling., Comment: Published in: 2023 IEEE International Conference on Cluster Computing (CLUSTER)

We report low-temperature powder X-ray and neutron diffraction studies on breathing pyrochlore magnets Cu$M$Cr$_{4}$S$_{8}$ ($M$ = Al, Ga), which undergo a magnetic transition at $T_{\rm N} \approx$ 21 and 31 K for {$M$ = Al and Ga, respectively. X-ray diffraction reveals that the magnetic transition accompanies a structural transition from cubic $F{\overline 4}3m$ to polar orthorhombic $Imm2$ symmetry for both the compounds, with larger distortion observed for $M$ = Ga at low temperatures. Neutron diffraction reveals incommensurate magnetic modulation ${\mathbf Q} = (q_{\rm IC}, 0.5, 0)$ in the orthorhombic setting, where $q_{\rm IC} \approx$ 0.39 and 0.31 for $M$ = Al and Ga, respectively. Our magnetic-structure analysis suggests cycloid-type magnetic order but not proper-screw type for both the compounds. We find strong correlation between the local spin configuration and Cr-Cr bond lengths, indicating that the spin-lattice coupling as well as the magnetic frustration play an important role in determining the ground state. Cu$M$Cr$_{4}$S$_{8}$ potentially offers a platform to explore magnetoelectric effects arising from the helimagnet driven electric polarity., Comment: 8 pages, 6 figures, 5 tables

CuAl2-type transition-metal (Tr) zirconides are superconductor family, and the Tr-site element substitution largely modifies its transition temperature (Tc). Here, we synthesized polycrystalline samples of Fe1-xNixZr2 by arc melting. From magnetic susceptibility measurements, bulk superconductivity was observed for 0.4 < x < 0.8, and the highest Tc of 2.8 K was observed for x = 0.6. Specific heat measurements were also performed, bulk superconductivity was observed for 0.4 < x < 0.8, and the highest Tc of 2.6 K was observed for x = 0.6. The obtained superconductivity phase diagram exhibits dome-shaped trend, which is similar to unconventional superconductors, where magnetic fluctuations are essential for superconductivity. In addition, from the c/a lattice constant ratio analysis, we show the possible relationship between the suppression of bulk superconductivity in the Ni-rich compositions and a collapsed tetragonal transition., Comment: 16 pages, 8 figures, SI

CPU-GPU heterogeneous architectures are now commonly used in a wide variety of computing systems from mobile devices to supercomputers. Maximizing the throughput for multi-programmed workloads on such systems is indispensable as one single program typically cannot fully exploit all available resources. At the same time, power consumption is a key issue and often requires optimizing power allocations to the CPU and GPU while enforcing a total power constraint, in particular when the power/thermal requirements are strict. The result is a system-wide optimization problem with several knobs. In particular we focus on (1) co-scheduling decisions, i.e., selecting programs to co-locate in a space sharing manner; (2) resource partitioning on both CPUs and GPUs; and (3) power capping on both CPUs and GPUs. We solve this problem using predictive performance modeling using machine learning in order to coordinately optimize the above knob setups. Our experiential results using a real system show that our approach achieves up to 67% of speedup compared to a time-sharing-based scheduling with a naive power capping that evenly distributes power budgets across components.

Recent High-Performance Computing (HPC) systems are facing important challenges, such as massive power consumption, while at the same time significantly under-utilized system resources. Given the power consumption trends, future systems will be deployed in an over-provisioned manner where more resources are installed than they can afford to power simultaneously. In such a scenario, maximizing resource utilization and energy efficiency, while keeping a given power constraint, is pivotal. Driven by this observation, in this position paper we first highlight the recent trends of resource management techniques, with a particular focus on malleability support (i.e., dynamically scaling resource allocations/requirements for a job), co-scheduling (i.e., co-locating multiple jobs within a node), and power management. Second, we consider putting them together, assess their relationships/synergies, and discuss the functionality requirements in each software component for future over-provisioned and power-constrained HPC systems. Third, we briefly introduce our ongoing efforts on the integration of software tools, which will ultimately lead to the convergence of malleability and power management, as it is designed in the HPC PowerStack initiative.

CPU-GPU heterogeneous systems are now commonly used in HPC (High-Performance Computing). However, improving the utilization and energy-efficiency of such systems is still one of the most critical issues. As one single program typically cannot fully utilize all resources within a node/chip, co-scheduling (or co-locating) multiple programs with complementary resource requirements is a promising solution. Meanwhile, as power consumption has become the first-class design constraint for HPC systems, such co-scheduling techniques should be well-tailored for power-constrained environments. To this end, the industry recently started supporting hardware-level resource partitioning features on modern GPUs for realizing efficient co-scheduling, which can operate with existing power capping features. For example, NVidia's MIG (Multi-Instance GPU) partitions one single GPU into multiple instances at the granularity of a GPC (Graphics Processing Cluster). In this paper, we explicitly target the combination of hardware-level GPU partitioning features and power capping for power-constrained HPC systems. We provide a systematic methodology to optimize the combination of chip partitioning, job allocations, as well as power capping based on our scalability/interference modeling while taking a variety of aspects into account, such as compute/memory intensity and utilization in heterogeneous computational resources (e.g., Tensor Cores). The experimental result indicates that our approach is successful in selecting a near optimal combination across multiple different workloads.

Magnetic flux trapping in field-cooled (FC) Sn-Pb solders has been recently studied because of the observation of nonvolatile magneto-thermal switching [H. Arima et al., Commun. Mater. 5, 34 (2024)] and anomalous magnetic field-temperature (H-T) phase diagrams [T. Murakami et al., AIP Adv. 13, 125008 (2023)]. In this paper, we investigate the origin of the anomalously low specific heat (C) in Sn10-Pb90 and Sn45-Pb55 solders after FC at H = 1500 Oe. We show that the FC solders exhibit self-heating possibly caused by the flux flow during the reduction of trapped fluxes when heating the sample during the C measurements. The T dependence of T rise clearly exhibits unexpectedly large values when the low-C states are observed. In addition, the cause of the transition-like behavior in C-T of FC solders are explained by local heating during H control and flux-jump phenomena., Comment: 19 pages, 7 figures, Supplemental data. arXiv admin note: text overlap with arXiv:2403.01719

We report the magnetic, magnetoelastic, and magnetotransport properties of single crystals of polar magnets $R$AuGe ($R=$ Dy, Ho, and Gd), grown by Au-Ge self-flux. Magnetization and magnetostriction measurements reveal multi-step metamagnetic transitions for the $c$-axis magnetic field ($H\parallel c$) for DyAuGe and HoAuGe, suggesting magnetic frustration in the triangular lattice of $R$ ions. The magnetic phase diagrams have clarified a close connection between the magnetoelastic property and the emergence of the intermediate metamagentic phase. The magnetic-field dependence of the resistivity and Hall resistivity reveal the semimetallic transport dominated by hole-type carriers, consistent with the behavior in a nonmagnetic analogue YAuGe. We also identify a signature of an anomalous Hall effect (AHE) proportional to the field-induced magnetization in $R=$ Dy, Ho, and Gd. GdAuGe shows magnetic and transport behavior as reported in a previous study using Bi-flux grown single crystals, while the self-flux grown crystal shows larger magnetoresistance ($\sim$ 345\%, at 1.8 K and 9 T) due to higher hole-type carrier mobility ($\sim$ 6400 cm$^2$/Vs). Using the two-band model analysis considering the mobility change during the magnetization process, we extract the anomalous Hall conductivity: $\sim 1200$ S/cm and $\sim 530$ S/cm for $R=$ Dy and Ho, respectively, at 1.8 K with 9 T for $H\parallel c$. The magnitude of conductivity suggests a contribution of intrinsic origin, possibly related to the Berry curvature in the electron bands induced by the time-reversal symmetry breaking and the polar lattice., Comment: 11 pages, 9 figures

The development of Large Language Models (LLMs) in various languages has been advancing, but the combination of non-English languages with domain-specific contexts remains underexplored. This paper presents our findings from training and evaluating a Japanese business domain-specific LLM designed to better understand business-related documents, such as the news on current affairs, technical reports, and patents. Additionally, LLMs in this domain require regular updates to incorporate the most recent knowledge. Therefore, we also report our findings from the first experiments and evaluations involving updates to this LLM using the latest article data, which is an important problem setting that has not been addressed in previous research. From our experiments on a newly created benchmark dataset for question answering in the target domain, we found that (1) our pretrained model improves QA accuracy without losing general knowledge, and (2) a proper mixture of the latest and older texts in the training data for the update is necessary. Our pretrained model and business domain benchmark are publicly available to support further studies., Comment: Accepted at PACLIC 38