Vectorizing optimization method based on MIC (Many Integrated Core) architecture processor

An optimization method, processor technology, applied in the field of scientific computing, computer high-performance computing

Active Publication Date: 2013-12-11
LANGCHAO ELECTRONIC INFORMATION IND CO LTD
View PDF4 Cites 25 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] With the promotion of MIC architecture processors, its powerful SIMD extension technology will be widely used, which not only provides new ways to solve problems and improve performance for high-performance computing, but also brings a new problem - how to quickly, Efficiently implement reliable vectorized parallel processing to fully unleash the computing potential of MIC processors? This is the real challenge facing software engineers

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
  • Vectorizing optimization method based on MIC (Many Integrated Core) architecture processor
  • Vectorizing optimization method based on MIC (Many Integrated Core) architecture processor
  • Vectorizing optimization method based on MIC (Many Integrated Core) architecture processor

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0073] The purpose of the present invention is to provide a vectorization optimization method based on MIC processor.

[0074] 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 conjunction with the accompanying drawings and embodiments.

[0075] First, briefly introduce the principle of vectorization processing and the architecture of the vector processing unit VPU (Vector Process Unit) of the MIC processor. The VPU of the MIC processor core supports KCi vector instructions with a bit width of 512bit, and supports multiple processing modes such as 16*32bit or 8*64bit, that is, the vectorization width is 8 or 16. 512 bits are equivalent to 16 single-precision floating-point data The length of the single-precision floating-point data vectorization processing diagram is as follows figure 1 Shown:

[0076] For example, the vector addition operation C[0~15]=A[0~15]+B[0~15] ...

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 provides a vectorizing optimization method based on an MIC architecture processor, which relates to three main steps, i.e. data dependence analysis for an algorithm, vectorizing adjustment and optimization of the algorithm and vectorizing compilation. The specific content includes data dependence analysis for the algorithm, the vectorizing optimization and adjustment of the algorithm, an automatic vectorization technique of a compiler, a user-intervened vectorizing optimization method and the like. The method provided by the invention is applicable to the software optimization of an MIC architecture processor platform, and can direct software developers to rapidly and efficiently carry out the vectorizing optimization of existing software, particularly core algorithms, within the shortest development cycles and with the lowest costs, consequently, the computing resources of vector processors can be utilized by software to the max, the running time of software can be shortened to the max, the hardware resource utilization rate can be remarkably increased, and the computing efficiency and overall performance of software can be increased.

Description

technical field [0001] The invention relates to the field of computer high-performance computing and the field of scientific computing, in particular to a vectorization optimization method based on an MIC architecture processor. Background technique [0002] Since Intel integrated MMX on the Pentium processor in 1996, more and more general-purpose processors have integrated SIMD (Single Instruction Multiple Data, Single Instruction Multiple Data) hardware extensions. processor, called a vector processor. The application of vector processors is also becoming more and more extensive, from the initial multimedia application to various application fields, especially in the field of high-performance computing, massive data and large-scale parallel processing requirements pose severe challenges to the computing power of processors, vectorization Parallel processing can effectively improve parallel processing efficiency and computing density, improve hardware resource utilization,...

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): G06F17/16
Inventor 吴庆
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