Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Heterogeneous cluster and task processing method and apparatus

A technology of heterogeneous clusters and tasks, applied in multi-programming devices, electrical digital data processing, physical implementation, etc., can solve the problems of increasing the speedup ratio, the proportion of time-consuming, and the large amount of synchronous communication data, so as to improve the execution efficiency. Effect

Active Publication Date: 2018-03-06
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF8 Cites 25 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] At present, it is common to use CPU clusters to accelerate deep learning applications. However, using CPU clusters to accelerate, although the acceleration ratio can be increased by increasing the cluster size, in the field of deep learning, as the cluster size increases, synchronous communication is required. The amount of data becomes larger, the proportion of time-consuming becomes heavier, and the acceleration ratio is limited

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
  • Heterogeneous cluster and task processing method and apparatus
  • Heterogeneous cluster and task processing method and apparatus
  • Heterogeneous cluster and task processing method and apparatus

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0050] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of this application.

[0051] figure 1 It is a schematic structural diagram of the heterogeneous cluster described in the present invention, such as figure 1 As shown, it includes multiple computing nodes, cluster resource management modules, Ethernet switches and IB (InfiniBand) network switches: Among them,

[0052] Computing nodes are used for deep learning applications; each ...

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 heterogeneous cluster and a task processing method and apparatus. The heterogeneous cluster comprises multiple calculation nodes, a cluster resource management module, an Ethernet exchanger and an IB network exchanger, wherein the calculation nodes are used for performing parallel calculation; each calculation node comprises a GPU and a CPU; the cluster resource management module is used for allocating the calculation nodes in the cluster and GPU resources on the calculation nodes; the calculation nodes are connected with the cluster resource management module throughthe Ethernet exchanger; and the calculation nodes are connected through the IB network exchanger. According to the scheme, the heterogeneous cluster can be used for performing the parallel calculation, so that the acceleration ratio is increased and the executive efficiency of a deep learning task is improved.

Description

【Technical field】 [0001] The invention relates to computer application technology, in particular to a heterogeneous cluster and task processing method and device. 【Background technique】 [0002] With the development of big data and deep learning technology, it is more and more valued by various technology companies to train and learn massive data through deep learning methods, and finally learn a set of accurate cognitive models. A more complex and powerful in-depth model can deeply reveal the complex and rich information carried in massive data, and make more accurate predictions for future or unknown events. Deep learning applications include speech recognition, image recognition, natural language processing, search advertisement CTR estimation, etc. These applications require a huge amount of computation. In the process of using deep learning for model training, a large number of computing tasks become the bottleneck restricting model training. [0003] If the traditiona...

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): G06F9/50G06N3/063
CPCG06F9/5027G06N3/063
Inventor 温圣召周汉清刘传秀张家军
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com