arXiv subjects

Kai Wu

Publications and source records attributed to Kai Wu.

At least 163 records · Page 9Linked to original sources

arXiv2020

MPC-based Vibration Control and Energy Harvesting Using an Electromagnetic Vibration Absorber With Inertia Nonlinearity

Simultaneous vibration control and energy harvesting of vehicle suspensions has attracted great research interests over the past decades. However, existing frameworks tradeoff suspension performance for energy recovery and are only responsive to narrow-bandwidth vibrations. In this paper, a new energy-regenerative vibration absorber (ERVA) using a ball-screw mechanism is investigated. The ERVA system is based on a rotary electromagnetic generator with adjustable nonlinear rotational inertia which passively increases the moment of inertia as the vibration amplitude increases. This structure is effective for energy harvesting and vibration control without increasing the suspension size. Furthermore, a nonlinear model predictive controller (NMPC) is applied to the system for further performance enhancement where we exploit road profile information as a preview. The performance of NMPC-based ERVA is demonstrated in a number of simulations and superior performance is demonstrated.

eess.SY↗

arXiv2020

Magnetic Particle Spectroscopy: A Short Review of Applications

Magnetic particle spectroscopy (MPS), also called magnetization response spectroscopy, is a novel measurement tool derived from magnetic particle imaging (MPI). It can be interpreted as a zero-dimensional version of MPI scanner. MPS was primarily designed for characterizing superparamagnetic iron oxide nanoparticles (SPIONs) regarding their applicability for MPI. In recent years, it has evolved into an independent, versatile, highly sensitive, inexpensive platform for biological and biomedical assays, cell labeling and tracking, and blood analysis. MPS has also developed into an auxiliary tool for magnetic imaging and hyperthermia by providing high spatial and temporal mappings of temperature and viscosity. Furthermore, other MPS-based applications are being explored such as magnetic fingerprints for product tracking and identification in supply chains. There are a variety of novel MPS-based applications being reported and demonstrated by many groups. In this short review, we highlighted some of the representative applications based on MPS platform, thereby providing a roadmap of this technology and our insights for researchers in this area.

physics.app-ph↗

arXiv2020

Single-core Or Multi-core? A Mini Review on Magnetic Nanoparticles for Magnetic Particle Spectroscopy-based Bioassays

Magnetic particle spectroscopy (MPS) is a technology that derives from magnetic particle imaging (MPI) and thrives as a standalone platform for many biological and biomedical applications, benefiting from the facile preparation and chemical modification of magnetic nanoparticles (MNPs). In recent years, MPS has been reported in extensive literatures as a versatile platform for different bioassay purposes using artificially designed MNPs, where the MNPs serve as magnetic tracers, the surface functionalized reagents (e.g., antibodies, aptamers, peptides, etc.) and tiny probes capturing target analytes from biofluid samples. The biochemical complexes on MNP surfaces can be tailored for different bioassay requirements, while the design of MNPs are of less attention for MPS-based bioassays. For MNPs in most bioassay applications, superparamagnetism is prerequisite to avoid agglomerates and false magnetic signals. Single- and multi-core superparamagnetic nanoparticles (SPMNPs) are prevalently used in MPS-based bioassays. In this mini review, we compared the pros & cons of different MPS platforms realizing volumetric- and surface-based bioassays with single- and multi-core nanoparticles, respectively.

physics.app-ph↗

arXiv2020

Demystifying the Performance of HPC Scientific Applications on NVM-based Memory Systems

The emergence of high-density byte-addressable non-volatile memory (NVM) is promising to accelerate data- and compute-intensive applications. Current NVM technologies have lower performance than DRAM and, thus, are often paired with DRAM in a heterogeneous main memory. Recently, byte-addressable NVM hardware becomes available. This work provides a timely evaluation of representative HPC applications from the "Seven Dwarfs" on NVM-based main memory. Our results quantify the effectiveness of DRAM-cached-NVM for accelerating HPC applications and enabling large problems beyond the DRAM capacity. On uncached-NVM, HPC applications exhibit three tiers of performance sensitivity, i.e., insensitive, scaled, and bottlenecked. We identify write throttling and concurrency control as the priorities in optimizing applications. We highlight that concurrency change may have a diverging effect on read and write accesses in applications. Based on these findings, we explore two optimization approaches. First, we provide a prediction model that uses datasets from a small set of configurations to estimate performance at various concurrency and data sizes to avoid exhaustive search in the configuration space. Second, we demonstrate that write-aware data placement on uncached-NVM could achieve $2$x performance improvement with a 60% reduction in DRAM usage.

cs.DC↗

arXiv2019

UMap: Enabling Application-driven Optimizations for Page Management

Leadership supercomputers feature a diversity of storage, from node-local persistent memory and NVMe SSDs to network-interconnected flash memory and HDD. Memory mapping files on different tiers of storage provides a uniform interface in applications. However, system-wide services like mmap are optimized for generality and lack flexibility for enabling application-specific optimizations. In this work, we present Umap to enable user-space page management that can be easily adapted to access patterns in applications and storage characteristics. Umap uses the userfaultfd mechanism to handle page faults in multi-threaded applications efficiently. By providing a data object abstraction layer, Umap is extensible to support various backing stores. The design of Umap supports dynamic load balancing and I/O decoupling for scalable performance. Umap also uses application hints to improve the selection of caching, prefetching, and eviction policies. We evaluate Umap in five benchmarks and real applications on two systems. Our results show that leveraging application knowledge for page management could substantially improve performance. On average, Umap achieved 1.25 to 2.5 times improvement using the adapted configurations compared to the system service.

cs.DC↗

arXiv2019

Irregularly Shaped γ'-Fe4N Nanoparticles for Hyperthermia Treatment and T2 Contrast-Enhanced Magnetic Resonance Imaging with Minimum Dose

Magnetic nanoparticles (MNPs) have been extensively used in drug/gene delivery, hyperthermia therapy, magnetic particle imaging (MPI), magnetic resonance imaging (MRI), magnetic bioassays, etc. With proper surface chemical modifications, physicochemically stable and non-toxic MNPs are emerging contrast agents and tracers for in vivo MRI and MPI applications. Herein, we report the high magnetic moment, irregularly shaped γ'-Fe4N nanoparticles for enhanced hyperthermia therapy and T2 contrast agent for MRI application. The static and dynamic magnetic properties of γ'-Fe4N nanoparticles are characterized by vibrating sample magnetometer (VSM) and magnetic particle spectroscopy (MPS) systems, respectively. Compared to the γ-Fe2O3 nanoparticles, γ'-Fe4N show at least 3 times higher saturation magnetization (in emu/g), which, as a result, gives rise to the stronger dynamic magnetic responses as proved in the MPS measurement results. In addition, γ'-Fe4N nanoparticles are functionalized with oleic acid layer by a wet mechanical milling process, the morphologies of as-milled nanoparticles are characterized by transmission electron microscopy (TEM), dynamic light scattering (DLS) and nanoparticle tracking analyzer (NTA). We report that with proper surface chemical modification and tuning on morphologies, γ'-Fe4N nanoparticles could be used as tiny heating sources for hyperthermia and contrast agents for MRI applications with minimum dose.

physics.app-ph↗

arXiv2019

Detection of magnetic nanoparticles (MNPs) using spin current nano-oscillator (SCNO) biosensor: A frequency-based rapid, ultra-sensitive, magnetic bioassay

This Letter is a micromagnetic simulation-based study on the GHz-frequency ferromagnetic resonances for the detection of magnetic nanoparticles (MNPs) using spin current nano-oscillator (SCNO) operating in precession mode as a spintronic biosensor. The magnetic stray fields from the MNPs in an antibody-antigen-MNP complex on the SCNO surface modify the ferromagnetic resonance peaks and generate measurable resonance peak shifts. Moreover, our results strongly indicate the position-sensitive behavior of the SCNO biosensor and ways to eradicate this effect to facilitate better bio-sensing performance. Additionally, a study has been made on how nanoparticles with different sizes can alter the SCNO device performance. This simulation-based study on the SCNO device shows a promise of frequency-based nano-biosensor with a sensitivity of detecting even a single MNP, even in presence of thermal noise.

physics.app-ph↗

arXiv2019

Sentinel: Runtime Data Management on Heterogeneous Main MemorySystems for Deep Learning

Software-managed heterogeneous memory (HM) provides a promising solution to increase memory capacity and cost efficiency. However, to release the performance potential of HM, we face a problem of data management. Given an application with various execution phases and each with possibly distinct working sets, we must move data between memory components of HM to optimize performance. The deep neural network (DNN), as a common workload on data centers, imposes great challenges on data management on HM. This workload often employs a task dataflow execution model, and is featured with a large amount of small data objects and fine-grained operations (tasks). This execution model imposes challenges on memory profiling and efficient data migration. We present Sentinel, a runtime system that automatically optimizes data migration (i.e., data management) on HM to achieve performance similar to that on the fast memory-only system with a much smaller capacity of fast memory. To achieve this,Sentinel exploits domain knowledge about deep learning to adopt a custom approach for data management. Sentinel leverages workload repeatability to break the dilemma between profiling accuracy and overhead; It enables profiling and data migration at the granularity of data objects (not pages), by controlling memory allocation. This method bridges the semantic gap between operating system and applications. By associating data objects with the DNN topology, Sentinel avoids unnecessary data movement and proactively triggers data movement. Using only 20% of peak memory consumption of DNN models as fast memory size, Sentinel achieves the same or comparable performance (at most 8% performance difference) to that of the fast memory-only system on common DNN models; Sentinel also consistently outperforms a state-of-the-art solution by 18%.

cs.PF↗

arXiv2019

Magnetic Nanoparticle Relaxation Dynamics-based Magnetic Particle Spectroscopy (MPS) for Rapid and Wash-free Molecular Sensing

Magnetic nanoparticles (MNPs) have been extensively used as contrasts and tracers for bioimaging, heating sources for tumor therapy, carriers for controlled drug delivery, and labels for magnetic immunoassays. Here, we describe a MNP relaxation dynamics-based magnetic particle spectroscopy (MPS) method for the quantitative detection of molecular biomarkers. In MPS measurements, the harmonics of oscillating MNPs are recorded and used as a metric for the freedom of rotational motion, which indicates the bound states of the MNPs. These harmonics can be collected from microgram quantities of iron oxide nanoparticles within 10 seconds. Using a streptavidin-biotin binding system, we demonstrate the feasibility of using MPS to sense these molecular interactions, showing this method is able to achieve rapid, wash-free bioassays, and is suitable for future point-of-care (POC), sensitive, and versatile diagnosis.

physics.app-ph↗

arXiv2019

Detection of Influenza A Virus Nucleoprotein Through the Self-Assembly of Nanoparticles in Magnetic Particle Spectroscopy-Based Bioassays: A Method for Rapid, Sensitive, and Wash-free Magnetic Immunoassays

Magnetic nanoparticles (MNPs) with proper surface functionalization have been extensively applied as labels for magnetic immunoassays, carriers for controlled drug/gene delivery, tracers and contrasts for magnetic imaging, etc. Here, we introduce a new biosensing scheme based on magnetic particle spectroscopy (MPS) and the self-assembly of MNPs to quantitatively detect H1N1 nucleoprotein molecules. MPS monitors the harmonics of oscillating MNPs as a metric for the freedom of rotational motion, thus indicating the bound states of MNPs. These harmonics can be readily collected from nanogram quantities of iron oxide nanoparticles within 10 s. H1N1 nucleoprotein molecule hosts multiple different epitopes that forms binding sites for many IgG polyclonal antibodies. Anchoring IgG polyclonal antibodies onto MNPs triggers the cross-linking between MNPs and H1N1 nucleoprotein molecules, thereby forming MNP self-assemblies. Using MPS and the self-assembly of MNPs, we achieved the sensitivity of 44 nM (442 pmole) for detecting H1N1 nucleoprotein. In addition, the morphologies and the hydrodynamic sizes of the MNP self-assemblies are characterized to verify the MPS results. Different MNP self-assembly models such as classical cluster, open ring tetramer, chain model as well as multimers (from dimer to pentamer) are proposed in this paper. Herein, we claim the feasibility of using MPS and the self-assembly of MNPs as a new biosensing scheme for detecting ultralow concentrations of target biomolecules, which can be employed as rapid, sensitive, and wash-free magnetic immunoassays.

physics.app-ph↗

arXiv2019

EasyCrash: Exploring Non-Volatility of Non-Volatile Memory for High Performance Computing Under Failures

Emerging non-volatile memory (NVM) is promising for building future HPC. Leveraging the non-volatility of NVM as main memory, we can restart the application using data objects remaining on NVM when the application crashes. This paper explores this solution to handle HPC under failures, based on the observation that many HPC applications have good enough intrinsic fault tolerance. To improve the possibility of successful recomputation with correct outcomes and ignorable performance loss, we introduce EasyCrash, a framework to decide how to selectively persist application data objects during application execution. Our evaluation shows that EasyCrash transforms 54% of crashes that cannot correctly recompute into the correct computation while incurring a negligible performance overhead (1.5% on average). Using EasyCrash and application intrinsic fault tolerance, 82% of crashes can successfully recompute. When EasyCrash is used with a traditional checkpoint scheme, it enables up to 24% improvement (15% on average) in system efficiency.

cs.PF↗

arXiv2019

Tunable Magnetic Domain Walls for Therapeutic Neuromodulation at Cellular Level: Stimulating Neurons Through Magnetic Nanowires

Cellular-level neuron stimulation has attracted much attention in the areas of prevention, diagnosis and treatment of neurological disorders. Herein, we propose a spintronic neurostimulator based on the domain wall movement inside stationary magnetic nanowires driven by the spin transfer torque. The electromotive forces generated by the domain wall motion can serve as highly localized stimulation signals for neuron cells. Our simulation results show that the induced electric field from the domain wall motion in permalloy nanowires can reach up to 14 V/m, which is well above the reported threshold stimulation signal for clinical applications. The proposed device operates on a current range of several uA which is 10^3 times lower compared to magnetic stimulation by microcoils. The duration and amplitude of the stimulating signal can be controlled by adjusting the applied current density, the geometry of the nanowire, and the magnetic properties of the nanowire material.

physics.med-ph↗

arXiv2019

Architecture-Aware, High Performance Transaction for Persistent Memory

Byte-addressable non-volatile main memory (NVM) demands transactional mechanisms to access and manipulate data on NVM atomically. Those transaction mechanisms often employ a logging mechanism (undo logging or redo logging). However, the logging mechanisms can bring large runtime overhead (8%-49% in our evaluation), and 41%-78% of the overhead attributes to the frequent cache-line flushing. Such large overhead significantly diminishes the performance benefits offered by NVM. In this paper, we introduce a new method to reduce the overhead of cache-line flushing for logging-based transactions. Different from the traditional method that works at the program level and leverages program semantics to reduce the logging overhead, we introduce architecture awareness. In particular, we do not flush certain cache blocks, as long as they are estimated to be eliminated out of the cache because of the hardware caching mechanism (e.g., the cache replacement algorithm). Furthermore, we coalesce those cache blocks with low dirtiness to improve the efficiency of cache-line flushing. We implement an architecture-aware, high performance transaction runtime system for persistent memory, Archapt. Our results show that comparing with an undo logging (PMDK) and a redo logging (Mnemosyne), Archapt reduces cache-line flushing by 66% and improves system throughput by 19% on average (42% at most). Our crash tests with four hardware caching policies show that Archapt provides a strong guarantee on crash consistency.

cs.DC↗

arXiv2019

Deterministic Field-free Switching of a Perpendicularly Magnetized Ferromagnetic Layer via the Joint Effects of Dzyaloshinskii-Moriya Interaction and Field-like Spin-orbit Torque: An Appraisal

Field-free switching of perpendicularly magnetized ferromagnetic layer by spin orbit torque (SOT) from the spin Hall effect (SHE) is of great interest in the applications of magnetic memory devices. In this paper, we investigate the deterministic SOT switching through the joint effects of Dzyaloshinskii-Moriya Interaction (DMI) and field-like torque (FLT) by micromagnetic simulations. We confirmed that within a certain range of DMI values and charge current densities, it is possible to deterministically switch the magnetization without the assistance external magnetic field. We show that the FLT could play an adverse role in blocking and slowing down the magnetization switching under certain cases of DMI and charge current-driven field-free switching. However, in other cases, FLT can assist DMI on the deterministic field-free SOT switching. In addition, it is found that FLT can effectively expand the current density window for a deterministic field-free SOT switching in the presence of DMI.

physics.app-ph↗

arXiv2019

Spin-orbit Torque and Spin Hall Effect-based Cellular Level Therapeutic Neuromodulators: Modulating Neuron Activities through Spintronic Nanodevices

Artificial modulation of a neuronal subset through ion channels activation can initiate firing patterns of an entire neural circuit in vivo. As nanovalves in the cell membrane, voltage-gated ion channels can be artificially controlled by the electric field gradient that caused by externally applied time varying magnetic fields. Herein, we theoretically investigate the feasibility of modulating neural activities by using magnetic spintronic nanostructures. An antiferromagnet/ferromagnet (AFM/FM) structure is explored as neuromodulator. For FM layer with perpendicular magnetization, stable bidirectional magnetization switching can be achieved by applying in-plane currents through AFM layer to induce the spin-orbit torque (SOT) due to the spin Hall effect (SHE). This Spin-orbit Torque Neurostimulator (SOTNS) utilizes in-plane charge current pulses to switch the magnetization in FM layer. The time changing magnetic stray field induces electric field that modulates the surrounding neurons. The Object Oriented Micromagnetic Framework (OOMMF) is used to calculate space and time dependent magnetic dynamics of SOTNS structure. The current driven magnetization dynamics in SOTNS has no mechanically moving parts. Furthermore, the size of SOTNS can be down to tens of nanometers, thus, arrays of SOTNSs could be fabricated, integrated together and patterned on a flexible substrate, which gives us much more flexible control of the neuromodulation with cellular resolution.

physics.app-ph↗

arXiv2019

Investigating the Effect of Magnetic Dipole-Dipole Interaction on Magnetic Particle Spectroscopy (MPS): Implications for Magnetic Nanoparticle-based Bioassays and Magnetic Particle Imaging (MPI)

Superparamagnetic iron oxide nanoparticles (SPIONs), with comparable size to biomolecules (such as proteins, nucleic acids, etc.) and unique magnetic properties, good biocompatibility, low toxicity, potent catalytic behavior, are promising candidates for many biomedical applications. There is one property present in most SPION systems, yet it has not been fully exploited, which is the dipole-dipole interaction (also called dipolar interaction) between the SPIONs. It is known that the magnetic dynamics of an ensemble of SPIONs are substantially influenced by the dipolar interactions. However, the exact way it affects the performance of magnetic particle-based bioassays and magnetic particle imaging (MPI) is still an open question. The purpose of this paper is to give a partial answer to this question. This is accomplished by numerical simulations on the dipolar interactions between two nearby SPIONs and experimental measurements on an ensemble of SPIONs using our lab-based magnetic particle spectroscopy (MPS) system. Our results show that even moderate changes in the SPION concentration may have substantial effects on the magnetic dynamics of the SPION system and the harmonic signal magnitudes can be increased or decreased by 60%, depending on the values of MPS system parameters.

physics.app-ph↗

arXiv2019

Tunable Magnetic Skyrmions in Ultrathin Magnetic Nanostructures for Cellular-Level Neurostimulation

In 2016, the Global Burden of Disease reported that neurological disorders were the principal cause of disability-adjusted life years (DALYs) and the second leading cause of deaths. Research in the last decade has pushed neuroscience to design and implement low-cost, efficient, implantable, flexible electrodes/probes and 3D arrays for neuron stimulation and sensing. Electrical arrays used in current CMOS-based technologies can be affected by the migration of cells (such as astrocytes) that attempt to seal off the electronic devices, causing increased impedance and alternations in the electric field. In this regard, magnetic nanodevices can be better candidates. A wide assortment of magnetic skyrmion-based device ideas and models have as of late been proposed featuring their potential applications. In this paper we propose a highly tunable skyrmion-based spintronic nanodevice for neuron stimulation. The effects of tunable material and magnetic properties specifically Dzyaloshinskii-Moriya interaction (DMI), perpendicular magnetic anisotropy (PMA) constant, number of skyrmions and device dimensions on stable skyrmion nucleation and smooth skyrmion dynamics in a magnetic ultra-thin film have been extensively studied. The aim of this study was to meet the standard therapeutic specifications of neuron stimulation, which is an electric field of about 10 mV/mm for a duration of 50 μs. From Faraday's Laws of Induction, skyrmion dynamics that generates an alternating magnetic flux density induces an electric field for a certain time duration in the ultra-thin film. The results of this work show that on tuning the skyrmion dynamics, the induced electric field can reach the standard value required for neurostimulation, thereby providing a strong futuristic possibility to exploit skyrmion-based spintronic nanodevices for neuron stimulation.

physics.app-ph↗

arXiv2018

Magnetic Nanoparticles in Nanomedicine

Nanomaterials, in addition to their small size, possess unique physicochemical properties that differ from the bulk materials, making them ideal for a host of novel applications. Magnetic nanoparticle (MNP) is one important class of nanomaterials that have been widely studied for their potential applications in nanomedicine. Due to the fact that MNPs can be detected and manipulated by remote magnetic fields, it opens a wide opportunity for them to be used in vivo. Nowadays, MNPs have been used for diverse applications including magnetic biosensing (diagnostics), magnetic imaging, magnetic separation, drug and gene delivery, and hyperthermia therapy, etc. This review aims to provide a comprehensive assessment of the state-of-the-art biological and biomedical applications of MNPs. In addition, the development of high magnetic moment MNPs with proper surface functionalization has progressed exponentially over the past decade. Herein, we also reviewed the recent advances in the synthesis and surface coating strategies of MNPs. This review is not only to provide in-depth insights into the different synthesis, biofunctionalization, biosensing, imaging, and therapy methods but also to give an overview of limitations and possibilities of each technology.

physics.app-ph↗