A High Performance Parallel and Heterogeneous Approach to Narrowband Beamforming

被引:1
|
作者
Sarofeen, Christian [1 ]
Gillett, Philip [2 ]
机构
[1] Naval Surface Warfare Ctr, Computat Anal & Design, Carderock Div, West Bethesda, MD 20817 USA
[2] Naval Surface Warfare Ctr, Hydroacoust & Propulsor Dev, Carderock Div, West Bethesda, MD 20817 USA
关键词
Beamforming; delay-sum beamforming; distributed computing; heterogeneous computing; hybrid parallel programming; ALGORITHM; FPGA; CUDA;
D O I
10.1109/TPDS.2015.2494038
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
This paper describes a high performing, hybrid parallel, and heterogeneous algorithmic approach to narrowband Delay-Sum Beamforming (DSB) in the frequency domain using a Just-In-Time Asynchronous Data Method (JIT-ADM) parallel pattern. JIT-ADM is a novel asynchronous parallel programming pattern that unifies various levels of asynchronous concurrency available with distributed heterogeneous computing. The computational performance of this DSB algorithm was analyzed on a 50 node Cray XC30 with a single 10-core Intel Xeon E5-2670 v2 and NVIDIA Tesla K20X general purpose Graphics Processing Unit (GPU) on each node. The algorithm exhibits well behaved weak scalability with 92.7 percent parallel efficiency at 50 nodes compared to maximum performance observed. It is also shown that the algorithm efficiently utilizes a large portion of the available hardware. During beamforming the GPU is utilized at 51.8 percent of its maximum double precision floating point throughput whereas a comparable Central Processing Unit (CPU) version utilizes 60.0 percent of its maximum expected floating point throughput. Across the weak scalability study, utilizing GPUs for processing, a 2-5x performance gain is achieved compared to using CPUs. A brief derivation and validation of the implemented DSB is also presented.
引用
收藏
页码:2196 / 2207
页数:12
相关论文
共 50 条
  • [41] Robust beamforming using a narrowband partially adaptive array
    Yang, H
    Ingram, MA
    NINTH IEEE SIGNAL PROCESSING WORKSHOP ON STATISTICAL SIGNAL AND ARRAY PROCESSING, PROCEEDINGS, 1998, : 13 - 16
  • [42] Adaptive orientational beamforming techniques for narrowband interference rejection
    Han, Jiangyan
    Ng, Boon Poh
    Er, Meng Hwa
    SIGNAL PROCESSING, 2022, 196
  • [43] Fault-Tolerant Parallel Scheduling of Tasks on a Heterogeneous High-Performance Workstation Cluster
    Yu-Kwong Kwok
    The Journal of Supercomputing, 2001, 19 : 299 - 314
  • [44] High-performance Shallow Water Model for Use on Massively Parallel and Heterogeneous Computing Systems
    Chaplygin A.V.
    Gusev A.V.
    Diansky N.A.
    Supercomputing Frontiers and Innovations, 2021, 8 (04) : 74 - 93
  • [45] Fault-tolerant parallel scheduling of tasks on a heterogeneous high-performance workstation cluster
    Kwok, YK
    JOURNAL OF SUPERCOMPUTING, 2001, 19 (03): : 299 - 314
  • [46] Design and initial performance of a high-level unstructured mesh framework on heterogeneous parallel systems
    Mudalige, G. R.
    Giles, M. B.
    Thiyagalingam, J.
    Reguly, I. Z.
    Bertolli, C.
    Kelly, P. H. J.
    Trefethen, A. E.
    PARALLEL COMPUTING, 2013, 39 (11) : 669 - 692
  • [47] An Improved Approach to Robust Capon Beamforming with Enhanced Performance
    Huang, Huiping
    Liao, Bin
    Guo, Chongtao
    Huang, Lei
    2016 CIE INTERNATIONAL CONFERENCE ON RADAR (RADAR), 2016,
  • [48] Bound performance models of heterogeneous parallel processing systems
    Balsamo, S
    Donatiello, L
    Van Dijk, NM
    IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 1998, 9 (10) : 1041 - 1056
  • [49] Performance evaluation of parallel haemodynamic computations on heterogeneous clouds
    Bystrov O.
    Kačeniauskas A.
    Pacevič R.
    Starikovičius V.
    Maknickas A.
    Stupak E.
    Igumenov A.
    Computing and Informatics, 2021, 39 (04) : 695 - 723
  • [50] Load and performance balancing scheme for heterogeneous parallel processing
    Kim, TH
    COMPUTATIONAL AND INFORMATION SCIENCE, PROCEEDINGS, 2004, 3314 : 298 - 303