Dynamic load balancing method for CPU+GPU CPPC

A dynamic load, parallel computing technology, applied in the direction of resource allocation, multi-program device, etc., to achieve the effect of improving the adaptability of the platform

Inactive Publication Date: 2014-07-02
LANGCHAO ELECTRONIC INFORMATION IND CO LTD
View PDF4 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the important and difficult point of how to realize the load balancing of CPU and GPU computing equipment collaborative computing, the present invention proposes a dynamic load balancing scheme and software implementation method of CPU+GPU collaborative parallel computing, which is effective in occasions where traditional dynamic load balancing methods are difficult to play a role solved the problem

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
  • Dynamic load balancing method for CPU+GPU CPPC
  • Dynamic load balancing method for CPU+GPU CPPC

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0037] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be described in detail below in combination with the text description and examples.

[0038] In the CPU+GPU cooperative parallel computing dynamic load balancing method described in the present invention, the software functions and operating procedures are summarized:

[0039] 1) Main thread parameter preprocessing;

[0040] 2) The main thread scans the GPU device model and quantity, and creates a structured device attribute (Handle) for each GPU device;

[0041] 3) The main thread performs other initialization operations, such as creating a progress bar, creating an output data body control handle, etc.;

[0042] 4) The main thread determines the number of enabled GPU devices N and the number of CPU computing cores M according to the input parameters;

[0043] 5) Main thread initialization (device) control thread parameter array;

[0044] 6) T...

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 load balancing method for CPU+GPU CPPC. The method mainly comprises the steps that in a single node, a main thread processes input parameters and finishes the corresponding initialization operation, 1+N computing equipment control threads are established, CPU devices and N GPU devices in the node are respectively controlled, a CPU control thread establishes a plurality of parallel computing threads according to the computing nuclear number in the node, a dynamic load balancing scheme is adopted between the CPU devices and the GPU devices, and optimization of the parallel efficiency is achieved. The number of enabled GPU calculating devices and the number of CPU computing threads can be manually allocated according to the user demand. According to the dynamic load balancing method, a pure CPU architecture platform and a CPU+GPU mixed architecture framework platform are combined in the load balancing scheme, a work station system with the various different types of GPU devices is supported, and the platform adaptability, the parallel efficiency and the overall running performance of software are greatly improved.

Description

technical field [0001] The invention relates to the field of computer high-performance computing and scientific computing, and in particular to a CPU+GPU cooperative parallel computing dynamic load balancing method. Background technique [0002] Driven by new technologies, hardware systems are undergoing rapid evolution, showing better performance and lower prices, such as the development of traditional CPUs from single-core to multi-core to many-core. With the popularization of multi-core architecture processors, multi-threaded parallel processing of application software will gradually become the mainstream. In addition, in addition to traditional applications (graphics display, mostly used in games), GPU is increasingly used in mathematical calculations due to its super floating-point computing capabilities, and has gradually entered the mainstream of computing. High-level languages ​​and development tools for GPU programming have also begun to emerge in large numbers. ...

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): G06F9/50
Inventor 吴庆张广勇吴韶华王娅娟
Owner LANGCHAO ELECTRONIC INFORMATION IND CO LTD
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