A value-at-risk simulation dynamic task scheduling method based on collaborative computing

A technology of dynamic tasks and scheduling methods, applied in the field of high-performance computing, can solve problems such as low utilization of computing resources and uneven task distribution, and achieve the effect of maximizing computing efficiency and realizing dynamic load

Active Publication Date: 2019-03-26
GUANGDONG UNIV OF TECH
View PDF10 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003]In the implementation of parallelization technology, it is easy to cause low utilization of computing resources due to uneven task distribution. Considering the operating efficiency, a key issue is how to To achieve load balancing

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A value-at-risk simulation dynamic task scheduling method based on collaborative computing
  • A value-at-risk simulation dynamic task scheduling method based on collaborative computing
  • A value-at-risk simulation dynamic task scheduling method based on collaborative computing

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0053] Now it is set that a value-at-risk Monte Carlo simulation is expected to produce 1 million simulation results, and the median value is taken as the simulation result. One computing unit of the computing platform used contains 1 CPU and 3 MICs, and the number of MIC cores is 61. At this point, the estimated total number of simulations N 0 =1000000, the value order α=0.5, the standard size P of the segmented task package D =2×61=122, the maximum number of segments

[0054]

[0055] When the task starts, the computing node generates a global simulation task queue, which contains 1024 maximum segments, and the number of segment task packets for each segment is

[0056]

[0057] then the actual number of simulations

[0058] N=N P ×N D ×P D =999424

[0059] loss rate

[0060]

[0061] Assume that the computing framework is actually composed of 1 computing node, each computing node has 8 management nodes, and each management node has 8 core nodes. The specif...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a dynamic task scheduling method of value-at-risk simulation based on collaborative computing. Under a three-layer parallel computing framework composed of a core node, a management node and a computing node, the computing node partitions tasks in a Monte Carlo simulation process and establishes a task queue. The task queue is composed of a core node, a management node anda computing node. The computing task snatching branch is carried out by the corresponding Pthread of the computing device; At last, that computing node sorts the compute results of the simulation process in parallel, and then merges and sort them to lighten the load of the management process. The invention mainly adopts the divide-and-conquer method to carry out dynamic task scheduling according to different computing devices of CPU and MIC, and realizes dynamic load balancing in the computing process so as to maximize the computing efficiency.

Description

technical field [0001] The present invention relates to the field of high-performance computing, and more specifically, to a dynamic task scheduling method for value-at-risk simulation based on collaborative computing. Background technique [0002] The Monte Carlo simulation calculation method of value at risk simulates the fluctuation of asset risk factors by generating a sequence of random numbers with corresponding distribution. Due to the high calculation consumption of the Monte Carlo simulation calculation method of value at risk, parallel computing technology needs to be used for optimization. Considering the load problem of parallel computing, the effect of dynamic scheduling technology will largely affect the final calculation efficiency. Regarding the parallelization feasibility of the calculation process of Monte Carlo simulation, after analysis, there is no correlation between the kth simulation and the k+1th simulation in the simulation process, and each indepen...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06F9/48G06F9/50
CPCG06F9/4881G06F9/5083
Inventor 程良伦邓博研王卓薇
Owner GUANGDONG UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap