arXiv subjects

Namhoon Kim

Publications and source records attributed to Namhoon Kim.

14 recordsLinked to original sources

arXiv2026

Physics-informed AI Accelerated Retention Analysis of Ferroelectric Vertical NAND: From Day-Scale TCAD to Second-Scale Surrogate Model

Ferroelectric field-effect transistors (FeFET)-based vertical NAND (Fe-VNAND) has emerged as a promising candidate to overcome z-scaling limitations with lower programming voltages. However, the data retention of 3D Fe-VNAND is hindered by the complex interaction between charge detrapping and ferroelectric depolarization. Developing optimized device designs requires exploring an extensive parameter space, but the high computational cost of conventional Technology Computer-Aided Design (TCAD) tools makes such wide-scale optimization impractical. To overcome these simulation barriers, we present a Physics-Informed Neural Operator (PINO)-based AI surrogate model designed for high-efficiency prediction of threshold voltage (Vth) shifts and retention behavior. By embedding fundamental physical principles into the learning architecture, our PINO framework achieves a speedup exceeding 10000x compared to TCAD while maintaining physical accuracy. The resulting surrogate provides a physics-consistent data engine for compact model parameter extraction and look-up-table (LUT) generation, directly supporting reliability-aware SPICE simulation of Fe-VNAND. This study demonstrates the model's effectiveness on a single FeFET configuration, serving as a pathway toward modeling the retention loss mechanisms.

cs.LG↗

arXiv2026

3D Field of Junctions: A Noise-Robust, Training-Free Structural Prior for Volumetric Inverse Problems

Volume denoising is a foundational problem in computational imaging, as many 3D imaging inverse problems face high levels of measurement noise. Inspired by the strong 2D image denoising properties of Field of Junctions (ICCV 2021), we propose a novel, fully volumetric 3D Field of Junctions (3D FoJ) representation that optimizes a junction of 3D wedges that best explain each 3D patch of a full volume, while encouraging consistency between overlapping patches. In addition to direct volume denoising, we leverage our 3D FoJ representation as a structural prior that: (i) requires no training data, and thus precludes the risk of hallucination, (ii) preserves and enhances sharp edge and corner structures in 3D, even under low signal to noise ratio (SNR), and (iii) can be used as a drop-in denoising representation via projected or proximal gradient descent for any volumetric inverse problem with low SNR. We demonstrate successful volume reconstruction and denoising with 3D FoJ across three diverse 3D imaging tasks with low-SNR measurements: low-dose X-ray computed tomography (CT), cryogenic electron tomography (cryo-ET), and denoising point clouds such as those from lidar in adverse weather. Across these challenging low-SNR volumetric imaging problems, 3D FoJ outperforms the evaluated classical denoisers, untrained neural denoisers, and denoisers trained only on noisy examples. Code is available at https://github.com/voilalab/3D-Field-of-Junctions.

cs.CV↗

arXiv2026

Perfusion Imaging and Single Material Reconstruction in Polychromatic Photon Counting CT

Background: Perfusion computed tomography (CT) images the dynamics of a contrast agent through the body over time, and is one of the highest X-ray dose scans in medical imaging. Recently, a theoretically justified reconstruction algorithm based on a monotone variational inequality (VI) was proposed for single material polychromatic photon-counting CT, and showed promising early results at low-dose imaging. Purpose: We adapt this reconstruction algorithm for perfusion CT, to reconstruct the concentration map of the contrast agent while the static background tissue is assumed known; we call our method VI-PRISM (VI-based PeRfusion Imaging and Single Material reconstruction). We evaluate its potential for dose-reduced perfusion CT, using a digital phantom with water and iodine of varying concentration. Methods: Simulated iodine concentrations range from 0.05 to 2.5 mg/ml. The simulated X-ray source emits photons up to 100 keV, with average intensity ranging from $10^5$ down to $10^2$ photons per detector element. The number of tomographic projections was varied from 984 down to 8 to characterize the tradeoff in photon allocation between views and intensity. Results: We compare VI-PRISM against filtered back-projection (FBP), and find that VI-PRISM recovers iodine concentration with error below 0.4 mg/ml at all source intensity levels tested. Even with a dose reduction between 10x and 100x compared to FBP, VI-PRISM exhibits reconstruction quality on par with FBP. Conclusion: Across all photon budgets and angular sampling densities tested, VI-PRISM achieved consistently lower RMSE, reduced noise, and higher SNR compared to filtered back-projection. Even in extremely photon-limited and sparsely sampled regimes, VI-PRISM recovered iodine concentrations with errors below 0.4 mg/ml, showing that VI-PRISM can support accurate and dose-efficient perfusion imaging in photon-counting CT.

physics.med-ph↗

arXiv2025

Short-range order influences H distribution in Fe-Ni-Cr austenitic stainless steels

Hydrogen embrittlement (HE) in austenitic stainless steels is advanced by hydrogen enhanced localized plasticity (HELP), typically accompanied by a transition from homogeneous to localized slip. Short-range order (SRO) in face-centered cubic (FCC) alloys is known to promote slip planarity, and recent studies suggest that H may amplify this localization behavior linked to inherent SRO. However, the manner in which the introduction of H affects SRO properties and, conversely, the manner that pre-existing SRO may affect H behavior, are not fully understood. In this work, a spin cluster expansion model combined with Monte Carlo simulation is employed to study the interplay between H and SRO in Fe-Ni-Cr alloys. Chemical order is quantified using Warren-Cowley SRO parameters, and the model predictions are validated against experimental data. We find that the presence of H only slightly alters the intrinsic ordering preference of the Fe-Ni-Cr alloys. As temperature decreases and the alloy evolves from disordered to ordered thermodynamic states, distinct H-metal correlations emerge. In particular, H-Ni and H-Cr pairs exhibit stronger ordering tendencies than H-Fe pairs, suggesting a selective affinity of H for certain atomic environments. On the other hand, we also find that compared to random alloys, when pre-existing SRO is present, it significantly affects the resulting H distribution by promoting local H enrichment in SRO domains. Such SRO-driven local H accumulation may facilitate slip localization and contribute to the early onset of embrittlement. These findings provide thermodynamic and structural insights into the interaction between H and SRO in austenitic stainless steels, highlighting possible implications on how the interaction between HELP and SRO brings about hydrogen embrittlement in austenitic stainless steels.

cond-mat.mtrl-sci↗

arXiv2025

Towards Distribution-Shift Uncertainty Estimation for Inverse Problems with Generative Priors

Generative models have shown strong potential as data-driven priors for solving inverse problems such as reconstructing medical images from undersampled measurements. While these priors improve reconstruction quality with fewer measurements, they risk hallucinating features when test images lie outside the training distribution. Existing uncertainty quantification methods in this setting (i) require an in-distribution calibration dataset, which may not be available, (ii) provide heuristic rather than statistical estimates, or (iii) quantify uncertainty from model capacity or limited measurements rather than distribution shift. We propose an instance-level, calibration-free uncertainty indicator that is sensitive to distribution shift, requires no knowledge of the training distribution, and incurs no retraining cost. Our key hypothesis is that reconstructions of in-distribution images remain stable under random measurement variations, while reconstructions of out-of-distribution (OOD) images exhibit greater instability. We use this stability as a proxy for detecting distribution shift. Our proposed OOD indicator is efficiently computable for any computational imaging inverse problem; we demonstrate it on tomographic reconstruction of MNIST digits, where a learned proximal network trained only on digit "0" is evaluated on all ten digits. Reconstructions of OOD digits show higher variability and correspondingly higher reconstruction error, validating this indicator. These results suggest a deployment strategy that pairs generative priors with lightweight guardrails, enabling aggressive measurement reduction for in-distribution cases while automatically warning when priors are applied out of distribution.

cs.CV↗

arXiv2025

Nitrogen-related short-range order in Fe-Ni-Cr austenitic stainless steels: first principles and cluster expansion study

Nitrogen (N) is a key alloying element that enhances the performance of Fe-Ni-Cr austenitic stainless steels, improving austenite stability, corrosion resistance, and yield strength. However, the role of N in modifying chemical ordering, particularly short-range order (SRO) and long-range order (LRO), is complex due to the multi-sublattice nature and magnetic interactions in these alloys. In this work, we combine first-principles calculations with the spin cluster expansion (spin CE) method to systematically investigate the effects of N on chemical ordering in Fe-Ni-Cr alloys. Our atomistic models confirm a strong affinity between N and Cr, which drives the formation of N-Cr SRO and, at higher N concentrations, stabilizes M4N-type ordered phases (M = metal). Monte Carlo simulations reveal that low N concentrations promote local N-Cr or N-N SRO, while increasing N content leads to the emergence of Cr-and N-rich LRO structures. We also show that the presence of N suppresses intrinsic Fe-Cr and Ni-Cr SRO by competing with these interactions, particularly at high concentrations. The impact of Cr content on ordering diminishes as N approaches its solubility limit. These findings are consistent with experimental observations in high-N austenitic steels. Finally, we discuss the influence of kinetic and magnetic effects on SRO evolution in high-N alloys. This study provides a comprehensive framework for understanding N-driven chemical ordering and offers insights into microstructural changes during nitriding processes.

cond-mat.mtrl-sci↗

arXiv2025

CLAMM: a spin CLuster expansion--Monte Carlo toolkit for Alloys and Magnetic Materials

Finite-temperature magnetism gives rise to many phenomena in alloy materials, such as magnetic phase transformations, short or medium range order in magnetic alloys, spin waves, critical phenomena, and the magnetocaloric effect. Lattice models, such as the Ising, Potts, cluster expansion, and magnetic cluster expansion models, are powerful tools for studying complex magnetic alloys and compounds. In this paper we introduce CLAMM, which is a new open source toolkit for developing custom lattice models from density functional theory (DFT) data sets. The toolkit is comprised of three main components. The first component is CLAMM_Prep, a python tool that converts data sets consisting of the Vienna Ab-initio Simulation Package (VASP) DFT simulations into a compact format. The second component, CLAMM_Fit, is also python-based and uses the compact data set to parameterize a lattice model, chosen from a set of available options (cluster expansion, Ising, and others). The third component is CLAMM_MC, which is a C++ Monte Carlo solver for generating ensembles of configurations, accounting for both magnetic and alloy configurational entropies, at different temperatures. These ensembles and their analysis can be used for simulating phase transformations and constructing phase diagrams. The code can also be used for generating special quasi-random structures and structures with user-defined short-range order. This document provides a comprehensive overview of each CLAMM tool in order to demonstrate CLAMM's potential for the computational materials community.

cond-mat.mtrl-sci↗

arXiv2025

Grids Often Outperform Implicit Neural Representations at Compressing Dense Signals

Implicit Neural Representations (INRs) have recently shown impressive results, but their fundamental capacity, implicit biases, and scaling behavior remain poorly understood. We investigate the performance of diverse INRs across a suite of 2D and 3D real and synthetic signals with varying effective bandwidth, as well as both overfitting and generalization tasks including tomography, super-resolution, and denoising. By stratifying performance according to model size as well as signal type and bandwidth, our results shed light on how different INR and grid representations allocate their capacity. We find that, for many tasks involving dense signals, a simple regularized grid with interpolation trains faster and to higher or comparable quality than any INR with the same number of parameters. We also find limited settings -- namely fitting binary signals such as shape contours -- where INRs outperform grids, to guide future development and use of INRs towards the most advantageous applications.

eess.IV↗

arXiv2024

First-principles and cluster expansion study of the effect of magnetism on short-range order in Fe-Ni-Cr austenitic stainless steels

Short-range order (SRO) alters the mechanical properties of technologically relevant structural materials such as medium/high entropy alloys and austenitic stainless steels. In this study, we present a generalized spin cluster expansion (CE) model and show that magnetism is a primary factor influencing the level of SRO present in austenitic Fe-Ni-Cr alloys. The spin CE consists of a chemical cluster expansion combined with an Ising model for Fe-Ni-Cr alloys. It explicitly accounts for local magnetic exchange interactions, thereby capturing the effects of finite temperature magnetism on SRO. Model parameters are obtained by fitting to a first-principles data set comprising both chemically and magnetically diverse FCC configurations. The magnitude of the magnetic exchange interactions are found to be comparable to the chemical interactions. Compared to a conventional implicit magnetism CE built from only magnetic ground state configurations, the spin CE shows improved performance on several experimental benchmarks over a broad spectrum of compositions, particularly at higher temperatures due to the explicit treatment of magnetic disorder. We find that SRO is strongly influenced by alloy Cr content, since Cr atoms prefer to align antiferromagnetically with nearest neighbors but become magnetically frustrated with increasing Cr concentration. We predict that increasing the Cr concentration in typical austenitic stainless steels promotes the formation of SRO and increases order-disorder transition temperatures. This study underscores the significance of considering magnetic interactions explicitly when exploring the thermodynamic properties of complex transition metal alloys. It also highlights guidelines for customizing SRO through adjustments of alloy composition.

cond-mat.mtrl-sci↗

arXiv2021

Mithril: Cooperative Row Hammer Protection on Commodity DRAM Leveraging Managed Refresh

Since its public introduction in the mid-2010s, the Row Hammer (RH) phenomenon has drawn significant attention from the research community due to its security implications. Although many RH-protection schemes have been proposed by processor vendors, DRAM manufacturers, and academia, they still have shortcomings. Solutions implemented in the memory controller (MC) incur increasingly higher costs due to their conservative design for the worst case in terms of the number of DRAM banks and RH threshold to support. Meanwhile, DRAM-side implementation either has a limited time margin for RH-protection measures or requires extensive modifications to the standard DRAM interface. Recently, a new command for RH-protection has been introduced in the DDR5/LPDDR5 standards, referred to as refresh management (RFM). RFM enables the separation of the tasks for RHprotection to both MC and DRAM by having the former generate an RFM command at a specific activation frequency and the latter take proper RH-protection measures within a given time window. Although promising, no existing study presents and analyzes RFM-based solutions for RH-protection. In this paper, we propose Mithril, the first RFM interfacecompatible, DRAM-MC cooperative RH-protection scheme providing deterministic protection guarantees. Mithril has minimal energy overheads for common use cases without adversarial memory access patterns. We also introduce Mithril+, an optional extension to provide minimal performance overheads at the expense of a tiny modification to the MC, while utilizing existing DRAM commands.

cs.CR↗

arXiv2020

Applying GPGPU to Recurrent Neural Network Language Model based Fast Network Search in the Real-Time LVCSR

Recurrent Neural Network Language Models (RNNLMs) have started to be used in various fields of speech recognition due to their outstanding performance. However, the high computational complexity of RNNLMs has been a hurdle in applying the RNNLM to a real-time Large Vocabulary Continuous Speech Recognition (LVCSR). In order to accelerate the speed of RNNLM-based network searches during decoding, we apply the General Purpose Graphic Processing Units (GPGPUs). This paper proposes a novel method of applying GPGPUs to RNNLM-based graph traversals. We have achieved our goal by reducing redundant computations on CPUs and amount of transfer between GPGPUs and CPUs. The proposed approach was evaluated on both WSJ corpus and in-house data. Experiments shows that the proposed approach achieves the real-time speed in various circumstances while maintaining the Word Error Rate (WER) to be relatively 10% lower than that of n-gram models.

cs.CL↗

arXiv2020

HEAAN Demystified: Accelerating Fully Homomorphic Encryption Through Architecture-centric Analysis and Optimization

Homomorphic Encryption (HE) draws a significant attention as a privacy-preserving way for cloud computing because it allows computation on encrypted messages called ciphertexts. Among numerous HE schemes proposed, HE for Arithmetic of Approximate Numbers (HEAAN) is rapidly gaining popularity across a wide range of applications because it supports messages that can tolerate approximate computation with no limit on the number of arithmetic operations applicable to the corresponding ciphertexts. A critical shortcoming of HE is the high computation complexity of ciphertext arithmetic; especially, HE multiplication (HE Mul) is more than 10,000 times slower than the corresponding multiplication between unencrypted messages. This leads to a large body of HE acceleration studies, including ones exploiting FPGAs; however, those did not conduct a rigorous analysis of computational complexity and data access patterns of HE Mul. Moreover, the proposals mostly focused on designs with small parameter sizes, making it difficult to accurately estimate their performance in conducting a series of complex arithmetic operations. In this paper, we first describe how HE Mul of HEAAN is performed in a manner friendly to computer architects. Then we conduct a disciplined analysis on its computational and memory access characteristics, through which we (1) extract parallelism in the key functions composing HE Mul and (2) demonstrate how to effectively map the parallelism to the popular parallel processing platforms, multicore CPUs and GPUs, by applying a series of optimization techniques such as transposing matrices and pinning data to threads. This leads to the performance improvement of HE Mul on a CPU and a GPU by 42.9x and 134.1x, respectively, over the single-thread reference HEAAN running on a CPU. The conducted analysis and optimization would set a new foundation for future HE acceleration research.

cs.DC↗

arXiv2018

Accelerating recurrent neural network language model based online speech recognition system

This paper presents methods to accelerate recurrent neural network based language models (RNNLMs) for online speech recognition systems. Firstly, a lossy compression of the past hidden layer outputs (history vector) with caching is introduced in order to reduce the number of LM queries. Next, RNNLM computations are deployed in a CPU-GPU hybrid manner, which computes each layer of the model on a more advantageous platform. The added overhead by data exchanges between CPU and GPU is compensated through a frame-wise batching strategy. The performance of the proposed methods evaluated on LibriSpeech test sets indicates that the reduction in history vector precision improves the average recognition speed by 1.23 times with minimum degradation in accuracy. On the other hand, the CPU-GPU hybrid parallelization enables RNNLM based real-time recognition with a four times improvement in speed.

cs.CL↗

arXiv2016

Screened-exchange density functional theory description of the electronic structure and phase stability of the chalcopyrite materials AgInSe$_2$ and AuInSe$_2$

We present a systematic assessment of the structural properties, the electronic density of states, the charge densities, and the phase stabilities of AgInSe$_2$ and AuInSe$_2$ using screened exchange hybrid density functional theory, and compare their properties to those of CuInSe$_2$. For AgInSe$_2$, hybrid density functional theory properly captures several experimentally measured properties, including the increase in the band gap and the change in the direction of the lattice distortion parameter $u$ in comparison to CuInSe$_2$. While the electronic properties of AuInSe$_2$ have not yet been experimentally characterized, we predict it to be a small gap ($\approx 0.15$ eV) semiconductor. We also present the phase stability of AgInSe$_2$ and AuInSe$_2$ according to screened-exchange density functional theory, and compare the results to predictions from conventional density functional theory, results tabulated from several online materials data repositories, and experiment (when available). In comparison to conventional density functional theory, the hybrid functional predicts phase stabilities of AgInSe$_2$ in better agreement with experiment: discrepancies in the calculated formation enthalpies are reduced by approximately a factor of three, from $\approx$ 0.20 eV/atom to $\approx$ 0.07 eV/atom, similar to the improvement observed for CuInSe$_2$. We further predict that AuInSe$_2$ is not a stable phase, and can only be present under non-equilibrium conditions.

cond-mat.mtrl-sci↗