用GPU加速信号形状与探测器分辨率随能量变化的卷积计算

乔颢; 王思广

doi:10.11889/j.0253-3219.2018.hjs.41.080404

您当前的位置：

首页 >

文章列表页 >

用GPU加速信号形状与探测器分辨率随能量变化的卷积计算

核电子学与仪器 | 更新时间：2021-05-19

- 用GPU加速信号形状与探测器分辨率随能量变化的卷积计算
- Apply GPU to accelerate the convolution computation of signal shape and detector resolution changed with energy
- 核技术 2018年41卷第8期文章编号：080404
- 作者机构：
  
  北京大学物理学院核物理与核技术国家重点实验室及高能物理中心北京 100871
- 作者简介：
  
  [ "乔颢, 男, 1993年出生, 2015年毕业于北京大学, 现为硕士研究生, 研究领域为高能物理实验模拟", "QIAO Hao, male, born in 1993, graduated from Peking University in 2015, master student, mainly engaged in the simulation in high energy physics" ]
  王思广, E-mail:siguang@pku.edu.cn WANG Siguang, E-mail:siguang@pku.edu.cn
- 基金信息：
  
  国家自然科学基金面上项目(10979010);中国科技部重点研发计划项目(2016YFA0400300)
- DOI：10.11889/j.0253-3219.2018.hjs.41.080404
  中图分类号： TL817.1
- 收稿日期：2018-03-28，
  
  修回日期：2018-04-12，
  
  纸质出版日期：2018-08
- 稿件说明：
移动端阅览
乔颢, 王思广. 用GPU加速信号形状与探测器分辨率随能量变化的卷积计算[J]. 核技术, 2018,41(8):080404

Hao QIAO, Siguang WANG. Apply GPU to accelerate the convolution computation of signal shape and detector resolution changed with energy[J]. Nuclear techniques, 2018, 41(8): 080404
乔颢, 王思广. 用GPU加速信号形状与探测器分辨率随能量变化的卷积计算[J]. 核技术, 2018,41(8):080404 DOI： 10.11889/j.0253-3219.2018.hjs.41.080404.

Hao QIAO, Siguang WANG. Apply GPU to accelerate the convolution computation of signal shape and detector resolution changed with energy[J]. Nuclear techniques, 2018, 41(8): 080404 DOI： 10.11889/j.0253-3219.2018.hjs.41.080404.

摘要

在形如中微子能谱分析中，因其本征能谱跨度较大，在积分描述所探测到的能谱时，探测器分辨率本身随能量的连续变化应该被考虑。对于该类能谱的拟合分析，传统的单靠中央处理器（Central Processing Unit，CPU）无法保障时间要求。本文利用图形处理器（Graphics Processing Unit，GPU）配合使用统一计算架构（Compute Unified Device Architecture，CUDA）和RooFit计算框架来加速数值积分过程，测试显示：当计算节点数目较多，如10

个节点时，利用文中的方法GPU（型号：K80）相较于CPU（型号：E5-2603 v3）能够带来218倍的速度提升。所述方法可作为类似实验的参考。

Abstract

Background

The relationship of detector resolution changed with energy should be considered when the intrinsic spectra have wide range in experiments like neutrino energy measurement. The normal numerical integration method using center processing unit (CPU) does not meet the performance requirement.

Purpose

This study aims to find a method which can accelerate the convolution operation without losing convenience.

Methods

The graphics processing unit (GPU) and compute unified device architecture (CUDA) toolkit are employed to implement RooFit's probability density function (PDF) class and calculate the numerical integration parallelly.

Results

In the test server using E5-2603 v3 as CPU and K80 as GPU

this method improves the speed as high as 218 times when the integration nodes set to be 1 million.

Conclusion

Combining GPU with RooFit library can get a significant improvement in the speed without lose RooFit's convenience. The method could be an example for those experiments which need complicated convolution calculations.

关键词

Keywords

references

X Qian , P Vogel . Neutrino mass hierarchy . Progress in Particle and Nuclear Physics , 2015 . 83 1 - 30 . DOI: 10.1016/j.ppnp.2015.05.002 http://doi.org/10.1016/j.ppnp.2015.05.002 .

W R Leo . Techniques for nuclear and particle physics experiments:a how-to approach , : Berlin Springer Science & Business Media , 2012 . 109 - 110 .

Weisstein, Eric W. Convolution theorem[EB/OL]. From MathWorld——A Wolfram Web Resource. 2018-5-24[2018-6-2] . http://mathworld.wolfram.com/ConvolutionTheorem.html http://mathworld.wolfram.com/ConvolutionTheorem.html .

The Toolkit for data modeling with ROOT[CP/DK]. 2018. https://root.cern.ch/roofit https://root.cern.ch/roofit .

王思广 , 冒亚军 , 唐培家 . CMGA:一个利用钚的94~104 keV γ和X射线非破坏分析其同位素丰度的软件 . 核技术 , 2006 . 29 ( 4 ): 257 - 260 . http://d.old.wanfangdata.com.cn/Periodical/hjs200604005 .

Siguang WANG , Yajun MAO , Peijia TANG . CMGA:a code to assay abundances of plutonium isotopes non-destructively with gamma and X-rays in 94~104 keV energy range . Nuclear Techniques , 2006 . 29 ( 4 ): 257 - 260 . http://d.old.wanfangdata.com.cn/Periodical/hjs200604005 .

S M Abrarov , B M Quine . Efficient algorithmic implementation of the Voigt/complex error function based on exponential series approximation . Applied Mathematics and Computation , 2011 . 218 ( 5 ): 1894 - 1902 . DOI: 10.1016/j.amc.2011.06.072 http://doi.org/10.1016/j.amc.2011.06.072 .

周铁 , 徐树方 , 张平文 , . 计算方法 , : 北京清华大学出版社 , 2006 . 55 - 71 .

Tie ZHOU , Shufang XU , Pingwen ZHANG , . Numerical analysis , : Beijing Tsinghua University Press , 2006 . 55 - 71 .

I Bogaert . Iteration-free computation of Gauss——legendre quadrature nodes and weights . SIAM Journal on Scientific Computing , 2014 . 36 ( 3 ): A1008 - 26 . DOI: 10.1137/140954969 http://doi.org/10.1137/140954969 .

ROOT data analyse framework[CP/DK]. 2018. https://root.cern.ch/ https://root.cern.ch/ .

B Angelucci , R Fantechi , G Lamanna , . The FPGA based trigger and data acquisition system for the CERN NA62 experiment . Journal of Instrumentation , 2014 . 9 ( 01 ): C01055 DOI: 10.1088/1748-0221/9/01/C01055 http://doi.org/10.1088/1748-0221/9/01/C01055 .

方民权 , 张卫民 . GPU编程与优化 , : 北京清华大学出版社 , 2016 . 10 - 16 .

Minquan FANG , Weimin ZHANG . GPU programming and code optimization , : Beijing Tsinghua University Press , 2016 . 10 - 16 .

CUDA zone[EB/OL]. 2018-6-2. https://developer.nvidia.com/cuda-zone https://developer.nvidia.com/cuda-zone .

OpenCL overview[EB/OL]. 2018-6-2. https://www.khronos.org/opencl/ https://www.khronos.org/opencl/ .

Cook Shano . CUDA并行程序设计, GPU编程指南 , : 北京机械工业出版社 , 2014 . 8 - 25, 61-70 .

Cook Shane . CUDA programming:a developer's guide to parallel computing with GPUs , : Beijing China Machine Press , 2016 . 8 - 25, 61-70 .

Thrust[CP/DK]. 2018. https://thrust.github.io/ https://thrust.github.io/ .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于CUDA的航空γ能谱数据小波降噪并行加速算法

基于数值积分的核截面在线生成方法研究

基于GPU并行计算实现快速显微CT重构