Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Budget power guidance-based high-energy-efficiency GPU (Graphics Processing Unit) cluster system scheduling algorithm

A technology of GPU cluster and scheduling method, applied in the field of high-efficiency GPU cluster system scheduling based on budget power guidance, can solve the problem of large energy consumption overhead of GPU cluster, and achieve the goal of improving energy use efficiency, high energy saving effect, and reducing energy consumption. Effect

Active Publication Date: 2015-05-20
TSINGHUA UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The computing power of GPU clusters is growing exponentially, but the ensuing problem is that the energy consumption of GPU clusters is also increasing

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
  • Budget power guidance-based high-energy-efficiency GPU (Graphics Processing Unit) cluster system scheduling algorithm
  • Budget power guidance-based high-energy-efficiency GPU (Graphics Processing Unit) cluster system scheduling algorithm
  • Budget power guidance-based high-energy-efficiency GPU (Graphics Processing Unit) cluster system scheduling algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.

[0033] The energy-efficient scheduling algorithm based on budget power guidance proposed by the present invention comprehensively uses four energy-saving strategies at different levels and different granularities, adopts a hierarchical scheduling method and adjusts the system power based on budget power guidance, so as to achieve the final High energy efficiency for long-term energy savings.

[0034] In order to achieve high-energy-efficiency scheduling, the present invention adopts four energy-saving strategies that help reduce system energy overhead, which are DRS energy-saving scheduling strategy, DVFS energy-saving scheduling strategy, critical path-based DAG task energy-saving scheduling strategy and β-migration energy-saving Scheduling strategy. In the present invention, according to the requirements of the budgeted power, these energy-...

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 budget power guidance-based high-energy-efficiency GPU (Graphics Processing Unit) cluster system scheduling algorithm, which comprises the following steps of: decomposing all tasks to be scheduled into basic tasks; dividing all the basic tasks into a plurality of independent task clusters; generating virtual nodes according to the task clusters to form one-to-one mapping relation between the task clusters and the virtual nodes; and finally performing many-to-many mapping on the virtual nodes to physical nodes to meet the requirement of dynamically allocating a proper processor to the tasks in the task clusters for execution of the tasks. According to the budget power guidance-based high-energy-efficiency GPU cluster system scheduling algorithm, the energy efficiency of a GPU cluster system can be effectively increased, so that an obvious energy-saving effect can be realized in the long term.

Description

technical field [0001] The invention relates to a GPU cluster system scheduling algorithm, in particular to an energy-efficient GPU cluster system scheduling method based on budget power guidance. Background technique [0002] The so-called GPU cluster refers to a high-performance computing cluster using a GPU (Graphics Processing Unit, graphics processing unit) as an accelerator. GPU has strong computing power, and at the same time has the advantages of easy deployment, low cost, and strong versatility. Typical GPU clusters include the "Tianhe-1A" and "Xingyun" supercomputers developed by my country. A GPU cluster generally includes thousands of homogeneous computing nodes, and each node is configured with multiple CPU processors and multiple GPU processors. Once a GPU cluster is officially put into use, it will run continuously for at least several years or even longer, and task scheduling is an essential system function accompanying the entire life cycle of a supercompu...

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 Patents(China)
IPC IPC(8): G06F9/50
CPCY02D10/00
Inventor 都志辉刘文杰刘渊杨全
Owner TSINGHUA UNIV
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
Eureka Blog
Learn More
PatSnap group products