摘要:
Sparse matrix-vector multiplication (SpMV) is an important issue in scientific computing and engineering applications. The performance of SpMV can be improved using parallel computing. The implementation and optimization of SpMV on GPU are research hotspots. Due to some irregularities of sparse matrices, the use of a single compression format is not satisfactory. The hybrid storage format can expand the range of adaptation of the compression algorithms. However, because of the imbalance of non-zero elements, the parallel computing capability of a GPU cannot be fully utilized. The parallel computing capability of a CPU is also rising due to increased number of cores in CPU. However, when a GPU is computing, the CPU controls the process instead of contributing to the computational work. It leads to under-utilization of the computing power of CPU. Due to the characteristics of the sparse matrices, the data can be split into two parts using the hybrid storage format to be allocated to CPU and GPU for simultaneous computing. In order to take full advantage of computing resources of CPU and GPU, the CPU-GPU heterogeneous computing model is adopted in this paper to improve the performance of SpMV. With analysis of the characteristics of CPU and GPU, an optimization strategy of sparse matrix partitioning using a distribution function is proposed to improve the computing performance of SpMV on the heterogeneous computing platform. The experimental results on two test machines demonstrate noticeable performance improvement. (C) 2017 Elsevier Inc. All rights reserved.
期刊:
Journal of Communications,2016年11(2):185-191 ISSN:1796-2021
通讯作者:
Shu, Wanneng(wangqianedu@163.com)
作者机构:
[Liu, Yang] College of Information Science and Engineering, Hunan City University, Yiyang, China;[Zhang, Chrish] Lawrence Berkeley National Laboratory, University of California, United States;[Shu, Wanneng] College of Computer Science, South-Central University for Nationalities, Wuhan, China
通讯机构:
College of Computer Science, South-Central University for Nationalities, Wuhan, China
作者机构:
[张弛; 王加阳] School of Information Science and Engineering, Central South University, Changsha, 410083, China;[贾丽媛; 张弛] School of Information Science and Engineering, Hunan City University, Yiyang, 413000, China
通讯机构:
School of Information Science and Engineering, Central South University, Changsha, China
期刊:
Multimedia Tools and Applications,2016年75(23):15601-15617 ISSN:1380-7501
通讯作者:
Zhang, Yudong
作者机构:
[Yang, Gelan] Hunan City Univ, Sch Informat Sci & Engn, Yiyang 413000, Peoples R China.;[Zhang, Yudong; Wang, Shuihua; Ji, Genlin] Nanjing Normal Univ, Sch Comp Sci & Technol, Nanjing 210023, Jiangsu, Peoples R China.;[Zhang, Yudong; Feng, Chunmei; Wang, Qiong; Yang, Jiquan] Jiangsu Key Lab 3D Printing Equipment & Mfg, Nanjing 210042, Jiangsu, Peoples R China.;[Dong, Zhengchao] Columbia Univ, Translat Imaging Div, New York, NY 10032 USA.;[Dong, Zhengchao] Columbia Univ, MRI Unit, New York, NY 10032 USA.
通讯机构:
[Zhang, Yudong] N;[Zhang, Yudong] J;Nanjing Normal Univ, Sch Comp Sci & Technol, Nanjing 210023, Jiangsu, Peoples R China.;Jiangsu Key Lab 3D Printing Equipment & Mfg, Nanjing 210042, Jiangsu, Peoples R China.
摘要:
It is very important to early detect abnormal brains, in order to save social and hospital resources. The wavelet-energy was a successful feature descriptor that achieved excellent performances in various applications; hence, we proposed a novel wavelet-energy based approach for automated classification of MR brain images as normal or abnormal. SVM was used as the classifier, and biogeography-based optimization (BBO) was introduced to optimize the weights of the SVM. The results based on a 5 x 5-fold cross validation showed the performance of the proposed BBO-KSVM was superior to BP-NN, KSVM, and PSO-KSVM in terms of sensitivity and accuracy. The study offered a new means to detect abnormal brains with excellent performance.
关键词:
Advanced Encryption Standard New Instruction;counter mode;encryption speed;encryption speedup;encryption time;OpenMP;parallel encryption;performance
摘要:
In the open environment of cloud computing, a large amount of user data needs to be encrypted/decrypted fast to maintain confidentiality and provide high quality of service. Advanced Encryption Standard (AES), the standard encryption algorithm, has better security and efficiency compared to its competitive algorithms, so it is widely used in cloud computing and other fields. However, the implementation of AES based on software still has the problem of low efficiency; whereas the implementation of AES based on hardware needs to purchase special purpose devices. Adopting the method of special instruction sets can resolve the above two drawbacks. Therefore, we propose a fast parallel cryptographic algorithm, NIPAES, which is based on the AES-NI (New Instructions) instruction set and CPU multiple cores. NIPAES makes use of the block property of AES and the parallel property of Counter (CTR) model, adopts OpenMP to evenly distribute workloads to each thread, which performs AES-NI instructions to complete encryption/decryption. Compared to CPU serial AES based on lookup tables, CPU parallel AES, and serial AES based on AES-NI, NIPAES has significant improvement on performance. The experimental results show that NIPAES achieves the average speedups of 3197.78x, 196.12x, and 7.71x, compared to the other aforementioned algorithms, respectively.
通讯机构:
[Tian, Ye] H;[Tian, Ye] U;Hunan City Univ, Sch Commun & Elect Engn, Yiyang 413000, Peoples R China.;Univ Ghent, Dept Informat Technol, Photon Res Grp, IMEC, B-9000 Ghent, Belgium.
关键词:
Meyer-Neldel rule;persistent photocurrent;trap-limited current
摘要:
A theoretical model is established to describe the emergence of the Meyer-Neldel rule (MNR) based on trap-limited current (TLC) theory. The model produces both MNR and anti-MNR behavior, and is available to various trap distributions. Moreover, TLC-based MNR is connected to persistent photocurrent (PPC) phenomenon. The information from MNR and PPC as well as their connection can be used to probe the distribution of the traps in materials.
作者机构:
[Shi, Zengliang; Zhu, Qiuxiang; Qin, Feifei; Zhu, Zhu; Lu, Junfeng; Xu, Chunxiang] Southeast Univ, Sch Biol Sci & Med Engn, State Key Lab Bioelect, Nanjing 210096, Jiangsu, Peoples R China.;[Zhu, Qiuxiang] Hunan City Univ, Coll Commun & Elect Engn, Yiyang 413000, Peoples R China.
通讯机构:
[Xu, Chunxiang] S;Southeast Univ, Sch Biol Sci & Med Engn, State Key Lab Bioelect, Nanjing 210096, Jiangsu, Peoples R China.
关键词:
Dual-band;Fabry-Perot lasing;ZnO microbelt;Vapor-phase transport
摘要:
Dual-band semiconductor microbelt lasing are promising for multifunctional applications ranging from optical communication to spectroscopy analysis. Here, we demonstrated a dual-band Fabry-Perot (F-P) lasing from both length and width directions in a single ZnO microbelt. The lasing performance, spectral variation and mode structure significantly depended on the cavity size, which corresponded to the length and width of the ZnO microbelts. The resonant process and mechanism were investigated systematically through the experimental analysis and numerically FDTD simulation. The results of the dual band F-P lasing modes and wide lasing wavelength are helpful to design the dual-wavelength electronic and optoelectronic devices. (C) 2016 Elsevier B.V. All rights reserved.
作者机构:
[阳同光] College of Information Science and Engineering, Hunan City University, Yiyang, 413000, China;[阳同光; 桂卫华] College of Information Science and Engineering, Central South University, Changsha, 410083, China
通讯机构:
College of Information Science and Engineering, Hunan City University, Yiyang, China
作者机构:
[燕锋; 丁兆明; 沈连丰; 朱亚萍] National Mobile Communications Research Laboratory, Southeast University, Nanjing, 210096, China;[邓曙光] College of Communications and Electronics Engineering, Hunan City University, Yiyang, 413000, China
期刊:
Modelling, Measurement and Control A,2016年89(1):188-204 ISSN:1259-5985
作者机构:
[Cui, Xianpu; Cui, Zhi] School of Communication and Electronic Engineering, Hunan City University China, No. 518, East Yingbin Road, Yiyang, China
关键词:
Fly back;Half-bridge resonant typology;High precision;Switching power supply
作者机构:
[刘琮; 杨格兰] School of Electronics and Information Engineering, Tongji University, Shanghai, 201804, China;[邓晓军] College of Computer and Communication, Hunan University of Technology, Zhuzhou, 412007, China;[杨格兰] School of Information Science and Engineering, Hunan City University, Yiang, 413000, China
通讯机构:
College of Computer and Communication, Hunan University of Technology, Zhuzhou, China
期刊:
材料导报,2016年30(5):148-151 and 155 ISSN:1005-023X
作者机构:
[张学军; 蒋练军; 张光富] School of Communication and Electronic Engineering, Hunan City University, Yiyang, 413000, China;[张光富] School of Physics and Electronics, Central South University, Changsha, 410083, China
摘要:
In this paper, the evolution of temporal soliton is investigated analytically when a laser pulse propagates in the inhomogeneous nonlinear medium with a Scarff II parity-time (PT)-symmetric potential. After a detailed analyzing the evolution of the intensity and pulse width (PW) of a temporal soliton, it is find that the chirped-free and chirped temporal soliton are stable when the dispersion coefficient is a periodic modulated function. When the dispersion coefficient are the constant and the exponential decreasing function, the chirped-free temporal soliton is stable, while the chirped temporal soliton is gradually compressed.
期刊:
Revista de la Facultad de Ingeniería Universidad Central de Venezuela,2016年31(11):144-156 ISSN:0798-4065
作者机构:
[He, Qian] College of Information Science and Engineering, Hunan City University, Yiyang, 413000, China;[He, Qian; Huang, Lihong] College of Mathematics and Econometrics, Hunan University, Changsha, 410082, China
摘要:
Recently, positron emission tomography (PET) has been widely used in medical image reconstruction. However, because of low tracer dosages and other reasons, the PET images are usually strongly polluted by noise, especially Poisson noise. The results of clinical diagnosis will be seriously affected by this noise. In order to suppress Poisson noise in reconstructed images, a new penalized maximum likelihood algorithm is proposed in this paper. It combines the Poisson-modified total variation model with the maximum likelihood expectation-maximization (MLEM) algorithm. Iterations of the proposed method can be divided into two steps: firstly, reconstructing image with the MLEM algorithm;secondly, suppressing Poison noise with the Poisson-modified total variation model. Experimental results demonstrate that the proposed method can effectively suppresses Poison noise in the PET images and is superior to many existing excellent algorithms.
作者机构:
[Li, Zimin] The Department of Art and Design, Hunan City University, Yiyang, 413000, Hunan, People's Republic of China;[Tian, Ye] School of Communication and Electronic Engineering, Hunan City University, Yiyang, 413000, Hunan, People's Republic of China