A device, method and instruction mapping method for executing instruction selection

A technology for executing instructions and instructions, which is applied in the field of instruction mapping of computers with data flow architecture, can solve problems such as low computing efficiency, and achieve the effect of improving computing efficiency and improving space-time utilization

Active Publication Date: 2021-07-20
上海睿伍科技有限公司
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0015] The purpose of the present invention is to solve the problem of low computational efficiency in the above-mentioned prior art, and propose a method and device for executing instruction selection, and at the same time, aiming at the selection method, an instruction mapping method based on the earliest executable time is proposed

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 device, method and instruction mapping method for executing instruction selection
  • A device, method and instruction mapping method for executing instruction selection
  • A device, method and instruction mapping method for executing instruction selection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0077] When the present invention studies the optimization of the instruction mapping algorithm on the multi-context data flow architecture computer, it is found that the defect of the prior art is caused by the blockage of the context pipeline during the execution of the data flow graph. As long as the blockage of the context is reduced, the The computing unit is effectively used in the subsequent execution, thereby improving the utilization rate of the computing unit and the throughput of the pipeline, and improving the computing efficiency. However, the random execution instruction selection method makes the instruction execution difficult to predict, and it is difficult to analyze and optimize the pipeline throughput.

[0078] If a fixed strategy can be adopted to select and execute instructions when the instruction is issued and selected, the behavior of the pipeline can be predicted, and then the pipeline can be analyzed and optimized.

[0079] In order to improve the ut...

Embodiment 2

[0149] In another specific embodiment, the present invention also provides a device for executing instruction selection, such as Figure 7 As shown, the device is composed of four instruction control cache components, several four-to-one selectors, and several judging components. Each instruction cache (that is, the instruction control cache component) stores the executable status of several instructions. The executable information of the same instruction in the four instruction caches is connected to the same 4-to-1 selector, and the 4-to-1 selector selects an executable instruction therefrom. Every four 4-to-1 selectors are sequentially connected to the four ports of a judging unit according to the order of the instructions, and the judging unit judges to select an executable command from the four 4-to-1 logics currently connected and execute it. It is sent to the output component, and the output signal terminates the operation of the subsequent components. If the instructi...

Embodiment 3

[0153] In another embodiment, the present invention also provides an instruction mapping method based on the earliest executable time, the flow chart of the method is as follows Figure 12 , the steps are as follows:

[0154] (1) Initialize the final execution time of all computing units to 0, and initialize the executable time of all macro instructions to 0;

[0155] (2) Sort all instructions according to the depth, and the instruction depth refers to the longest path length from the current instruction to the root node (start node);

[0156] (3) In order of depth, the instructions are mapped sequentially. For each instruction, traverse all computing units;

[0157] (4) For each computing unit, calculate the executable time of the current instruction mapped on the computing unit. In a specific implementation, the calculation method is: calculate the sum of the completion time of the first instruction of all predecessor macro-instructions of the current instruction and the ...

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 present invention provides an execution instruction selection device, method, and instruction mapping method. The execution instruction selection method includes: when there are multiple executable instructions on a computing unit at the same time, selecting the earlier of the multiple executable instructions The mapped instructions are executed; when there are multiple instructions with the same mapping sequence among the multiple executable instructions, one of the multiple instructions with the same mapping sequence is randomly selected for execution. This scheme effectively improves the calculation efficiency and improves the space-time utilization of the calculation unit.

Description

technical field [0001] The invention relates to the field of computer architecture, in particular to the field of instruction mapping methods for computers with data flow architecture. Background technique [0002] With the development of basic science, large-scale scientific applications have put forward new requirements for computing power. Because traditional architecture computers encounter bottlenecks such as heat dissipation, energy consumption, and technology, the data flow architecture has become a reliable choice for a new generation of high-performance computers with its advantages of high concurrency, simple control, and low energy consumption. [0003] Computers with data flow architecture are generally composed of two-dimensional computing unit arrays. The computing units are connected through a high-speed network. The calculation is completed by computing data in the computing unit and passing between computing units. Each computing unit is placed with one or m...

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): G06F9/30
CPCG06F9/3005
Inventor 高龑谭旭李文明马丽娜冯煜晶张浩
Owner 上海睿伍科技有限公司
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