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

Floating-point number conversion method and device

A conversion method and floating-point number technology, which is applied in the computer field and can solve the problems of low training efficiency of neural networks

Pending Publication Date: 2020-06-19
NANJING UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a floating-point number conversion method and device to solve the problem of low efficiency of neural network training caused by using single-precision floating-point numbers based on the IEEE 754 specification

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
  • Floating-point number conversion method and device
  • Floating-point number conversion method and device
  • Floating-point number conversion method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0078] figure 2 A flow chart of the floating-point number conversion method provided by Embodiment 1 of the present invention. Such as figure 2 As shown, a floating-point number conversion method provided by the embodiment of the present invention includes the following steps:

[0079] S101. Obtain the value of the first symbol segment, the value of the first exponent segment, and the value of the first mantissa segment in the first floating-point number respectively, the first floating-point number is a single-precision floating-point number in a normalized data format, that is, based on IEEE A normalized single-precision floating-point number of the 754 specification.

[0080] Generally, for a normalized single-precision floating-point number based on the IEEE 754 specification, the normalized single-precision floating-point number representation of the floating-point number is:

[0081] A=(-1) S ×2 21-127 ×1.F,

[0082] Among them, E1 is the value of the exponent segm...

Embodiment 2

[0099] In the floating-point number conversion method provided by Embodiment 2 of the present invention, on the basis of Embodiment 1 above, the value of the first index segment and the preset index bit width are used to determine the value of the organization segment and the second index segment The value of the step specific can include:

[0100] S201. Determine the value of the organization segment by using the value of the first index segment and the preset index bit width.

[0101] Specifically, the following formula is used in this embodiment to determine the value of the tissue segment:

[0102] r=[E / 2 es ],

[0103] Wherein, r represents the value of the organization segment, and when r is a non-integer value, the value of r is rounded down; E represents the value of the first index segment, and es represents the bit width of the preset index.

[0104] Taking the number whose true value is 0.125 as an example, it is expressed in the form of floating-point scientific...

Embodiment 3

[0113] In the floating-point conversion method provided by the third embodiment of the present invention, on the basis of the above-mentioned embodiment, the value of the second symbol segment, the value of the organization segment, the value of the second exponent segment and the The step of forming the second floating-point number in the form of binary code according to the preset total bit width of the value of the second mantissa segment specifically includes:

[0114] S301. Using the value of the organization segment, determine a binary code corresponding to the value of the organization segment.

[0115] For floating-point numbers in the Posit data format, the value of the organization segment r is floating. In data representation, the encoding of the organization segment r has two representations: one is continuous 1 and a subsequent 0, such as 111...0; the other is continuous 0 and a subsequent 1, such as 000... 1. For the real value r of the organizational segment, ...

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 floating-point number conversion method and a floating-point number conversion device, which can convert a single-precision floating-point number based on an IEEE 754 specification into a floating-point number in a posit data format, namely a second floating-point number. In the training process of a plurality of neural networks, the operation data approximately obeys normal distribution; data can be concentrated near 0 through transformation; however, the floating-point number of the posit data format in the invention can ensure the precision near 0 in the neural network training process; moreover, the preset total bit width of the floating-point number in the posit data format can be regulated and controlled, so that the data bit width can be reduced to a greatextent, resources required for storage and resources consumed in the read-write process are reduced, and the efficiency of neural network training is improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a floating-point number conversion method and device. Background technique [0002] Neural network is an algorithmic mathematical model that imitates the behavior characteristics of animal neural networks and performs distributed parallel information processing. This kind of network depends on the complexity of the system, and achieves the purpose of processing information by adjusting the interconnection relationship between a large number of internal nodes. In recent years, with the rapid development of deep learning technology, the training of neural networks has become common and important, and the speed and resource consumption of neural network training have also become important indicators for deep learning evaluation. [0003] In the previous neural network training process, most of the floating-point numbers used the normalized single-precision floating-point number f...

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): H03M7/24
CPCH03M7/24Y02D10/00
Inventor 王中风徐铭阳方超林军
Owner NANJING UNIV
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