Development of new materials in hard drive designs requires characterization of nanoscale materials through grain segmentation. The high-throughput quickly changing research environment makes zero-shot generalization an incredibly desirable feature. For this reason, we explore the application of Meta's Segment Anything Model (SAM) to this problem. We first analyze the out-of-the-box use of SAM. Then we discuss opportunities and strategies for improvement under the assumption of minimal labeled data availability. Out-of-the-box SAM shows promising accuracy at property distribution extraction. We are able to identify four potential areas for improvement and show preliminary gains in two of the four areas.

With the rapid development of AI technology in recent years, there have been many studies with deep learning models in soft sensing area. However, the models have become more complex, yet, the data sets remain limited: researchers are fitting million-parameter models with hundreds of data samples, which is insufficient to exercise the effectiveness of their models and thus often fail to perform when implemented in industrial applications. To solve this long-lasting problem, we are providing large scale, high dimensional time series manufacturing sensor data from Seagate Technology to the public. We demonstrate the challenges and effectiveness of modeling industrial big data by a Soft Sensing Transformer model on these data sets. Transformer is used because, it has outperformed state-of-the-art techniques in Natural Language Processing, and since then has also performed well in the direct application to computer vision without introduction of image-specific inductive biases. We observe the similarity of a sentence structure to the sensor readings and process the multi-variable sensor readings in a time series in a similar manner of sentences in natural language. The high-dimensional time-series data is formatted into the same shape of embedded sentences and fed into the transformer model. The results show that transformer model outperforms the benchmark models in soft sensing field based on auto-encoder and long short-term memory (LSTM) models. To the best of our knowledge, we are the first team in academia or industry to benchmark the performance of original transformer model with large-scale numerical soft sensing data.

In the era of big data, data-driven based classification has become an essential method in smart manufacturing to guide production and optimize inspection. The industrial data obtained in practice is usually time-series data collected by soft sensors, which are highly nonlinear, nonstationary, imbalanced, and noisy. Most existing soft-sensing machine learning models focus on capturing either intra-series temporal dependencies or pre-defined inter-series correlations, while ignoring the correlation between labels as each instance is associated with multiple labels simultaneously. In this paper, we propose a novel graph based soft-sensing neural network (GraSSNet) for multivariate time-series classification of noisy and highly-imbalanced soft-sensing data. The proposed GraSSNet is able to 1) capture the inter-series and intra-series dependencies jointly in the spectral domain; 2) exploit the label correlations by superimposing label graph that built from statistical co-occurrence information; 3) learn features with attention mechanism from both textual and numerical domain; and 4) leverage unlabeled data and mitigate data imbalance by semi-supervised learning. Comparative studies with other commonly used classifiers are carried out on Seagate soft sensing data, and the experimental results validate the competitive performance of our proposed method.

Striking a balance between protecting data privacy and enabling collaborative computation is a critical challenge for distributed machine learning. While privacy-preserving techniques for federated learning have been extensively developed, methods for scenarios involving bitwise operations, such as tree-based vertical federated learning (VFL), are still underexplored. Traditional mechanisms, including Shamir's secret sharing and multi-party computation (MPC), are not optimized for bitwise operations over binary data, particularly in settings where each participant holds a different part of the binary vector. This paper addresses the limitations of existing methods by proposing a novel binary multi-party computation (BiMPC) framework. The BiMPC mechanism facilitates privacy-preserving bitwise operations, with a particular focus on dot product computations of binary vectors, ensuring the privacy of each individual bit. The core of BiMPC is a novel approach called Dot Product via Modular Addition (DoMA), which uses regular and modular additions for efficient binary dot product calculation. To ensure privacy, BiMPC uses random masking in a higher field for linear computations and a three-party oblivious transfer (triot) protocol for non-linear binary operations. The privacy guarantees of the BiMPC framework are rigorously analyzed, demonstrating its efficiency and scalability in distributed settings.

NAND flash memory is ubiquitous in everyday life today because its capacity has continuously increased and cost has continuously decreased over decades. This positive growth is a result of two key trends: (1) effective process technology scaling; and (2) multi-level (e.g., MLC, TLC) cell data coding. Unfortunately, the reliability of raw data stored in flash memory has also continued to become more difficult to ensure, because these two trends lead to (1) fewer electrons in the flash memory cell floating gate to represent the data; and (2) larger cell-to-cell interference and disturbance effects. Without mitigation, worsening reliability can reduce the lifetime of NAND flash memory. As a result, flash memory controllers in solid-state drives (SSDs) have become much more sophisticated: they incorporate many effective techniques to ensure the correct interpretation of noisy data stored in flash memory cells. In this chapter, we review recent advances in SSD error characterization, mitigation, and data recovery techniques for reliability and lifetime improvement. We provide rigorous experimental data from state-of-the-art MLC and TLC NAND flash devices on various types of flash memory errors, to motivate the need for such techniques. Based on the understanding developed by the experimental characterization, we describe several mitigation and recovery techniques, including (1) cell-tocell interference mitigation; (2) optimal multi-level cell sensing; (3) error correction using state-of-the-art algorithms and methods; and (4) data recovery when error correction fails. We quantify the reliability improvement provided by each of these techniques. Looking forward, we briefly discuss how flash memory and these techniques could evolve into the future.

In this paper, we propose a new approach to construct a class of check-hybrid generalized low-density parity-check (CH-GLDPC) codes which are free of small trapping sets. The approach is based on converting some selected check nodes involving a trapping set into super checks corresponding to a 2-error correcting component code. Specifically, we follow two main purposes to construct the check-hybrid codes; first, based on the knowledge of the trapping sets of the global LDPC code, single parity checks are replaced by super checks to disable the trapping sets. We show that by converting specified single check nodes, denoted as critical checks, to super checks in a trapping set, the parallel bit flipping (PBF) decoder corrects the errors on a trapping set and hence eliminates the trapping set. The second purpose is to minimize the rate loss caused by replacing the super checks through finding the minimum number of such critical checks. We also present an algorithm to find critical checks in a trapping set of column-weight 3 LDPC code and then provide upper bounds on the minimum number of such critical checks such that the decoder corrects all error patterns on elementary trapping sets. Moreover, we provide a fixed set for a class of constructed check-hybrid codes. The guaranteed error correction capability of the CH-GLDPC codes is also studied. We show that a CH-GLDPC code in which each variable node is connected to 2 super checks corresponding to a 2-error correcting component code corrects up to 5 errors. The results are also extended to column-weight 4 LDPC codes. Finally, we investigate the eliminating of trapping sets of a column-weight 3 LDPC code using the Gallager B decoding algorithm and generalize the results obtained for the PBF for the Gallager B decoding algorithm.

Depth resolved study of structural and magnetic profiles of antiferromagnetic/ferromagnetic (AFM/FM) system upon annealing was performed in this work. We studied systems comprising of AFM IrMn and FM (Co, Fe, Co$_{70}$Fe$_{30}$) structures using polarized neutron and soft X-ray scattering, secondary neutral spectrometry, and magnetometry. Structural depth profiles obtained from neutron reflectometry indicate non-homogeneity of the AFM layer even before annealing, which is associated with the migration of manganese to the surface of the sample. Annealing of samples with CoFe and Co layers leads to a slight increase ($\sim$ 5 %) in the migration of manganese, which, however, does not lead to significant degradation of the exchange coupling at the AFM/FM interface. A significantly different picture was observed in the Fe/IrMn systems where a strong migration of iron into the AFM layer was observed upon annealing of the sample, leading to erosion of the magnetic profile, the formation of a non-magnetic alloy and degradation of the pinning strength. This study can be useful in the design of AF/FM systems in different spintronics devices, including HDD read heads, where thermal annealing is applied at different stages of the device manufacturing process.

This paper focuses on designing a privacy-preserving Machine Learning (ML) inference protocol for a hierarchical setup, where clients own/generate data, model owners (cloud servers) have a pre-trained ML model, and edge servers perform ML inference on clients' data using the cloud server's ML model. Our goal is to speed up ML inference while providing privacy to both data and the ML model. Our approach (i) uses model-distributed inference (model parallelization) at the edge servers and (ii) reduces the amount of communication to/from the cloud server. Our privacy-preserving hierarchical model-distributed inference, privateMDI design uses additive secret sharing and linearly homomorphic encryption to handle linear calculations in the ML inference, and garbled circuit and a novel three-party oblivious transfer are used to handle non-linear functions. privateMDI consists of offline and online phases. We designed these phases in a way that most of the data exchange is done in the offline phase while the communication overhead of the online phase is reduced. In particular, there is no communication to/from the cloud server in the online phase, and the amount of communication between the client and edge servers is minimized. The experimental results demonstrate that privateMDI significantly reduces the ML inference time as compared to the baselines.

We have used a plane-wave expansion method to theoretically study the far-field head-media optical interaction in HAMR. For the ASTC media stack specifically, we notice the outstanding sensitivity related to interlayer's optical thickness for media reflection and magnetic layer's light absorption. With 10-nm interlayer thickness change, the recording layer absorption can be changed by more than 25%. The 2-D results are found to correlate well with full 3-D model and magnetic recording tests on flyable disc with different interlayer thickness.

The development of permanent magnets containing less or no rare-earth elements is linked to profound knowledge of the coercivity mechanism. Prerequisites for a promising permanent magnet material are a high spontaneous magnetization and a sufficiently high magnetic anisotropy. In addition to the intrinsic magnetic properties the microstructure of the magnet plays a significant role in establishing coercivity. The influence of the microstructure on coercivity, remanence, and energy density product can be understood by {using} micromagnetic simulations. With advances in computer hardware and numerical methods, hysteresis curves of magnets can be computed quickly so that the simulations can readily provide guidance for the development of permanent magnets. The potential of rare-earth reduced and free permanent magnets is investigated using micromagnetic simulations. The results show excellent hard magnetic properties can be achieved in grain boundary engineered NdFeB, rare-earth magnets with a ThMn12 structure, Co-based nano-wires, and L10-FeNi provided that the magnet's microstructure is optimized.

Compared to planar (i.e., two-dimensional) NAND flash memory, 3D NAND flash memory uses a new flash cell design, and vertically stacks dozens of silicon layers in a single chip. This allows 3D NAND flash memory to increase storage density using a much less aggressive manufacturing process technology than planar NAND flash memory. The circuit-level and structural changes in 3D NAND flash memory significantly alter how different error sources affect the reliability of the memory. In this paper, through experimental characterization of real, state-of-the-art 3D NAND flash memory chips, we find that 3D NAND flash memory exhibits three new error sources that were not previously observed in planar NAND flash memory: (1) layer-to-layer process variation, where the average error rate of each 3D-stacked layer in a chip is significantly different; (2) early retention loss, a new phenomenon where the number of errors due to charge leakage increases quickly within several hours after programming; and (3) retention interference, a new phenomenon where the rate at which charge leaks from a flash cell is dependent on the data value stored in the neighboring cell. Based on our experimental results, we develop new analytical models of layer-to-layer process variation and retention loss in 3D NAND flash memory. Motivated by our new findings and models, we develop four new techniques to mitigate process variation and early retention loss in 3D NAND flash memory. These four techniques are complementary, and can be combined together to significantly improve flash memory reliability. Compared to a state-of-the-art baseline, our techniques, when combined, improve flash memory lifetime by 1.85x. Alternatively, if a NAND flash vendor wants to keep the lifetime of the 3D NAND flash memory device constant, our techniques reduce the storage overhead required to hold error correction information by 78.9%.

Recent research on CoPd alloys with perpendicular magnetic anisotropy (PMA) has suggested that they might be useful as the pinning layer in CoFeB/MgO-based perpendicular magnetic tunnel junctions (pMTJ's) for various spintronic applications such as spin-torque transfer random access memory (STT-RAM). We have previously studied the effect of seed layer and composition on the structure (by XRD, SEM, AFM and TEM) and performance (coercivity) of these CoPd films. These films do not switch coherently, so the coercivity is determined by the details of the switching mechanism, which was not studied in our previous paper. In the present paper, we show that information can be obtained about the switching mechanism from magnetic force microscopy (MFM) together with first order reversal curves (FORC), despite the fact that MFM can only be used at zero field. We find that these films switch by a mechanism of domain nucleation and dendritic growth into a labyrinthine structure, after which the unreversed domains gradually shrink to small dots and then disappear.

In this paper we apply an extended Landau-Lifschitz equation, as introduced by Ba\v{n}as et al. for the simulation of heat-assisted magnetic recording. This equation has similarities with the Landau-Lifshitz-Bloch equation. The Ba\v{n}as equation is supposed to be used in a continuum setting with sub-grain discretization by the finite-element method. Thus, local geometric features and nonuniform magnetic states during switching are taken into account. We implement the Ba\v{n}as model and test its capability for predicting the recording performance in a realistic recording scenario. By performing recording simulations on 100 media slabs with randomized granular structure and consecutive read back calculation, the write position shift and transition jitter for bit lengths of 10nm, 12nm, and 20nm are calculated.

Over the years, hardware trends have introduced various heterogeneous compute units while also bringing network and storage bandwidths within an order of magnitude of memory subsystems. In response, developers have used increasingly exotic solutions to extract more performance from hardware; typically relying on static, design-time partitioning of their programs which cannot keep pace with storage systems that are layering compute units throughout deepening hierarchies of storage devices. We argue that dynamic, just-in-time partitioning of computation offers a solution for emerging data-intensive systems to overcome ever-growing data sizes in the face of stalled CPU performance and memory bandwidth. In this paper, we describe our prototype computational storage system (CSS), Skytether, that adopts a database perspective to utilize computational storage drives (CSDs). We also present MSG Express, a data management system for single-cell gene expression data that sits on top of Skytether. We discuss four design principles that guide the design of our CSS: support scientific applications; maximize utilization of storage, network, and memory bandwidth; minimize data movement; and enable flexible program execution on autonomous CSDs. Skytether is designed for the extra layer of indirection that CSDs introduce to a storage system, using decomposable queries to take a new approach to computational storage that has been imagined but not yet explored. In this paper, we evaluate: partition strategies, the overhead of function execution, and the performance of selection and projection. We expected ~3-4x performance slowdown on the CSDs compared to a consumer-grade client CPU but we observe an unexpected slowdown of ~15x, however, our evaluation results help us set anchor points in the design space for developing a cost model for decomposable queries and partitioning data across many CSDs.

Magnetic droplets are nanoscale, non-topological, magnetodynamical solitons that can be nucleated in spin torque nano-oscillators (STNOs) or spin Hall nano-oscillators (SHNOs). All theoretical, numerical, and experimental droplet studies have so far focused on the free layer (FL), and any additional dynamics in the reference layer (RL) have been entirely ignored. Here we show, using all-perpendicular STNOs, that there is not only significant magnetodynamics in the RL, but the reference layer itself can host a droplet coexisting with the FL droplet. Both droplets are observed experimentally as stepwise changes and sharp peaks in the dc and differential resistance, respectively. Whereas the single FL droplet is highly stable, the coexistence state exhibits high-power broadband microwave noise. Micromagnetic simulations corroborate the experimental results and reveal a strong interaction between the droplets. Our demonstration of strongly interacting and closely spaced droplets offers a unique platform for fundamental studies of highly non-linear soliton pair dynamics.

Multi-level magnetic recording is a new concept for increasing the data storage capacity of hard disk drives. However, its implementation has been limited by a lack of suitable media capable of storing information at multiple levels. Herein, we overcome this problem by developing dual FePt-C nanogranular films separated by a Ru-C breaking layer with a cubic crystal structure. The FePt grains in the bottom and top layers of the developed media exhibited different effective magnetocrystalline anisotropies and Curie temperatures. The former is realized by different degrees of ordering in the L10-FePt grains, whereas the latter was attributed to the diffusion of Ru, thereby enabling separate magnetic recordings at each layer under different magnetic fields and temperatures. Furthermore, the magnetic measurements and heat-assisted magnetic recording simulations showed that these media enabled 3-level recording and could potentially be extended to 4-level recording, as the up-down and down-up states exhibited non-zero magnetization.

Over the last few decades, modern industrial processes have investigated several cost-effective methodologies to improve the productivity and yield of semiconductor manufacturing. While playing an essential role in facilitating real-time monitoring and control, the data-driven soft-sensors in industries have provided a competitive edge when augmented with deep learning approaches for wafer fault-diagnostics. Despite the success of deep learning methods across various domains, they tend to suffer from bad performance on multi-variate soft-sensing data domains. To mitigate this, we propose a soft-sensing ConFormer (CONvolutional transFORMER) for wafer fault-diagnostic classification task which primarily consists of multi-head convolution modules that reap the benefits of fast and light-weight operations of convolutions, and also the ability to learn the robust representations through multi-head design alike transformers. Another key issue is that traditional learning paradigms tend to suffer from low performance on noisy and highly-imbalanced soft-sensing data. To address this, we augment our soft-sensing ConFormer model with a curriculum learning-based loss function, which effectively learns easy samples in the early phase of training and difficult ones later. To further demonstrate the utility of our proposed architecture, we performed extensive experiments on various toolsets of Seagate Technology's wafer manufacturing process which are shared openly along with this work. To the best of our knowledge, this is the first time that curriculum learning-based soft-sensing ConFormer architecture has been proposed for soft-sensing data and our results show strong promise for future use in soft-sensing research domain.

Four different ways of obtaining low-density parity-check codes from expander graphs are considered. For each case, lower bounds on the minimum stopping set size and the minimum pseudocodeword weight of expander (LDPC) codes are derived. These bounds are compared with the known eigenvalue-based lower bounds on the minimum distance of expander codes. Furthermore, Tanner's parity-oriented eigenvalue lower bound on the minimum distance is generalized to yield a new lower bound on the minimum pseudocodeword weight. These bounds are useful in predicting the performance of LDPC codes under graph-based iterative decoding and linear programming decoding.

Using numerical and analytical micromagnetics we calculated the switching fields and energy barriers of the composite (exchange spring) magnetic recording media, which consist of layers with high and low magnetocrystalline anisotropy. We demonstrate that the ultimate potential of the composite media is realized if the interfacial domain wall fits inside the layers. The switching occurs via domain wall nucleation, compression in the applied field, de-pinning and propagation through the hard/soft interface. This domain wall assisted switching results in a significant reduction of the switching field without substantial decrease of the for thermal activation energy barrier. We demonstrate that the Domain Wall Assisted Magnetic Recording (DWAMR) offers up to a three-fold areal density gain over conventional single layer recording.