A5048606069- Advanced Data Storage Technologies
- Cloud Computing and Resource Management
- Distributed and Parallel Computing Systems
- Caching and Content Delivery
- Parallel Computing and Optimization Techniques
- Peer-to-Peer Network Technologies
- Distributed systems and fault tolerance
- Mechatronics Education and Applications
- Experimental Learning in Engineering
- Advanced Database Systems and Queries
- Data Management and Algorithms
- Engineering Education and Pedagogy
- Antenna Design and Optimization
- Engineering and Test Systems
- Advanced Computational Techniques and Applications
- Embedded Systems Design Techniques
- Advanced Semiconductor Detectors and Materials
- High-Voltage Power Transmission Systems
- Optical Systems and Laser Technology
- Power Systems Fault Detection
- Network Packet Processing and Optimization
- Machine Learning in Bioinformatics
- Evaluation Methods in Various Fields
- 3D Surveying and Cultural Heritage
- Full-Duplex Wireless Communications
Los Alamos National Laboratory
2022-2024
Shanghai Jiao Tong University
2011-2022
Carnegie Mellon University
2014-2021
Gannon University
2011-2020
PLA Army Engineering University
2019
Northwestern Polytechnical University
2012
Nanjing Institute of Technology
2009
Institute of Electronics
2006
Hua Yuan Group (China)
2001
California Digital Library
2000
The growing size of modern storage systems is expected to exceed billions objects, making metadata scalability critical overall performance. Many existing distributed file only focus on providing highly parallel fast access data, and lack a scalable service. In this paper, we introduce middleware design called Index FS that adds support such as PVFS, Lustre, HDFS for high-performance operations small files. uses table-based architecture incrementally partitions the namespace per-directory...
Modern key-value stores often use write-optimized indexes and compact in-memory to speed up read write performance. One popular index is the Log-structured merge-tree (LSM-tree) which provides indexed access write-intensive data. It has been increasingly used as a storage backbone for many services, including file system metadata management, graph processing engines, machine learning feature engines. Existing LSM-tree implementations exhibit high amplifications caused by compaction, lack...
High performance computing fault tolerance depends on scalable parallel file system performance. For more than a decade bandwidth has been available from the object storage systems that underlie modern systems, and recently we have seen demonstrations of metadata using dynamic partitioning namespace over multiple servers. But even these require significant numbers dedicated servers, some workloads still experience bottlenecks. We envision exascale do not any server machines. Instead job...
The rapid growth of cloud storage systems calls for fast and scalable namespace processing. While few commercial file offer anything better than federating individually non-scalable servers, a recent academic system, IndexFS, demonstrates processing based on client caching directory entries permissions (directory lookup state) with no per-client state in servers. In this paper we explore explicit replication all servers as an alternative to information clients. Both eliminate most repeated...
With object storage services becoming increasingly accepted as replacements for traditional file or block systems, it is important to effectively measure the performance of these services. Thus people can compare different solutions tune their systems better performance. However, little has been reported on this specific topic yet. To address problem, we present COSBench (Cloud Object Storage Benchmark), a benchmark tool that are currently working in Intel cloud In addition, paper, also...
With object storage systems being increasingly recognized as a preferred way to expose one's infrastructure the web, past few years have witnessed an explosion in acceptance of these systems. Unfortunately, proliferation available solutions and complexity each individual one, coupled with lack dedicated workload, makes it very challenging for one evaluate tune performance different To help address this problem, we present Cloud Object Storage Benchmark (COSBench). It is benchmark tool that...
Parallel file systems are often characterized by a layered architecture that decouples metadata management from I/O operations, allowing to facilitate fast concurrent access contents. However, intensive workloads still likely bottleneck at the system control plane due namespace synchronization, which taxes application performance through lock contention on directories, transaction serialization, and RPC overheads. In this paper, we propose client-driven architecture, BatchFS, is optimized...
Analysis of large-scale simulation output is a core element scientific inquiry, but analysis queries may experience significant I/O overhead when the data not structured for efficient retrieval. While in-situ processing allows improved time-to-insight many applications, scaling frameworks to hundreds thousands cores can be difficult in practice. The DeltaFS indexing new approach massive amounts achieve point and small-range queries. This paper describes challenges lessons learned this...
A number of challenges in implementing cloud technique related to further improving Web application performance and decreasing the cost. In order achieve high profits, cloud-based web providers must carefully balance resources dynamic workloads. However, this task is usually difficulty because complex nature most application. paper, we presented a predictive model analyze such applications determine when how much resource allocate each tier an addition, proposed new profit describe revenues...
Key–value (KV) software has proven useful to a wide variety of applications including analytics, time-series databases, and distributed file systems. To satisfy the requirements diverse workloads, KV stores have been carefully tailored best match performance characteristics underlying solid-state block devices. Emerging storage device is promising technology for both simplifying stack improving persistent storage-based applications. However, while providing fast, predictable put get...
Popular software key-value stores such as LevelDB and RocksDB are often tailored for efficient writing. Yet, they tend to also perform well on read operations. This is because while data initially stored in a format that favors writes, it later transformed by the DB background into better accommodates reads. Write-optimized can still block writes. happens when those workers cannot keep up with foreground insertion workload.This paper advocates hardware-accelerated store, enabling...
In this paper, we present a novel data acquisition scheme based on low-noise front-end readout ASIC and high speed ADC for PET imaging systems cadmium zinc telluride (CZT) detectors. A charge-sensitive amplifier(CSA), pulse shaper driving buffer are integrated each CZT detector pixel. The specifications of the dependent dimension detector. Eight shaping voltages sampled digitized. from is collected by programmable FPGA which can run an algorithm to calculate peak value shaped trigging timing...
In this paper we introduce the Indexed Massive Directory, a new technique for indexing data within DeltaFS. With its design as scalable, server-less file system HPC platforms, DeltaFS scales metadata performance with application scale. The Directory is novel extension to plane, enabling in-situ of massive amounts written single directory simultaneously, and in an arbitrarily large number files. We achieve through memory-efficient mechanism reordering data, log-structured storage layout pack...
Abstract INTEGRATED STEM-BASED PROJECTS TO INSPIRE K-12 STUDENTS PURSUE UNDERGRADUATE DEGREE PROGRAMS IN ELECTRICAL AND COMPUTER ENGINEERINGAbstractThis paper discusses how integrated electrical and computer engineering (ECE) projectswith science, technology, engineering, mathematics (STEM) components can inspirethe students to pursue the undergraduate degree programs in ECE. These projectsare presented through Engineering day events hosted at baccalaureatedegree institution by ECE...
Complex storage stacks providing data compression, indexing, and analytics help leverage the massive amounts of generated today to derive insights. It is challenging perform this computation, however, while fully utilizing underlying media. This because, servers with large core counts are widely available, single-core performance memory bandwidth per grow slower than count die. Computational offers a promising solution problem by dedicated compute resources along processing path. We present...
High-Performance Computing (HPC) is known for its use of massive concurrency. But it can be challenging a parallel filesystem's control plane to utilize cores when every client process must globally synchronize and serialize metadata mutations with those other clients. We present DeltaFS, new paradigm distributed filesystem metadata. DeltaFS allows jobs self-commit their namespace changes logs, avoiding the cost global synchronization. Followup selectively merge logs produced by previous as...
This paper discusses the work in progress to restructure Electronics curriculum Department of Electrical and Computer Engineering (ECE) order improve system integration learning experience gained by undergraduate students. The goals restructuring ECE are as follows: a) train prepare students design analyze complex electronic systems first at subsystem level before teaching electronics component level; b) strengthen infrastructure for with other courses such power through use integrated...
In order to optimize test flow after the default was modified by hand, a new software framework for radar fault isolation illustrated. This separated all mapping algorithms from graph so as modify and insert algorithm dynamically in testing process. Based on this framework, kind of optimization method proposed studied. By defining an objective function, we could evaluate candidate flows get optimizing flow. An example explained how search flows.
In this paper, a RF high power Doherty feed-forward amplifier using equal group-delay canceller is proposed which provide efficiency and linearity also with improved performance. The takes Freescale mrf21125S N-channel Enhancement-Mode Lateral MOSFETs as the main of amplifier. By carefully adjust gate bias phase offset line well through 3 dB hybrid canceller, at output 25 W performs 44 better adjacent channel leakage ratio (ACLR), 3.9% higher added (PAE) than conventional Class-AB