High-speed split multiply accumulator MAC apparatus

A multiply-accumulator and multiply-accumulate technology, which is applied in the field of high-speed split multiply-accumulator MAC devices, can solve problems such as low calculation parallelism and inability to meet multimedia streams, and achieve high calculation parallelism, fast operating frequency, and pipeline no depth effect

Inactive Publication Date: 2009-03-04
ZHEJIANG UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the calculation parallelism of existing multiply-accumulate MAC devices is not high, usually only one multiplication or multiply-accumulate operation can be performed in one cycle, which cannot meet the needs of multimedia streams
Moreover, high-speed processors require the multiply-accumulate unit to also have high-speed performance. In some designs, the depth of the pipeline is sacrificed in order to achieve high-speed requirements, so that it takes multiple cycles to output a result when starting the multiply-accumulate operation.

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
  • High-speed split multiply accumulator MAC apparatus
  • High-speed split multiply accumulator MAC apparatus
  • High-speed split multiply accumulator MAC apparatus

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0023] In conjunction with the accompanying drawings, the present invention will be described in detail below.

[0024] exist figure 1 In the example of the present invention, the high-speed split multiplier-accumulator MAC device is divided into four sequentially executed functional units, and each functional unit completes part of the functions of the MAC device.

[0025] I is a functional unit I, which is the data selection logic of the multiplier. It receives multi-bit multiplier and multiplicand data and granularity selection signal, and it selects appropriate data according to the granularity selection signal to output to four multipliers. Its specific logic is:

[0026] Set G_mul high: perform full word (32bit) operation.

[0027] The operands of multiplier 1 are: AH, BH

[0028] The operands of multiplier 2 are: AL, BL

[0029] The operands of multiplier 3 are: AH, BL

[0030] The operands of multiplier 4 are: AL, BH

[0031] G_mul set low: execute half word (16...

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

This invention discloses one high speed crack sum MAC device, which comprises input and output and four function units, wherein, the unit one is data selection unit; function unit two is semi-bit wide sum up device; function unit three comprise two parts of circuit with one for two and half mode sum up output circuit and two for whole word mode multiplier or sum up pre-process circuit; the function unit four is for whole mode multiplier or sum up output circuit with one as bit in selection structure composed of selector and one bit sum device.

Description

technical field [0001] The invention relates to a high-speed split multiply-accumulator MAC device for solving multiple multiply-accumulate algorithms. Background technique [0002] In digital signal processing applications, many operations, such as FIR, IIR, convolution algorithm, and FFT, etc., involve various modes of multiplication or multiply-accumulate operations, and the processing capability of the multiply-accumulate algorithm directly affects the performance of the processor. The multiplier and the multiplicand of the multiplication and accumulation operation are likely to be signed numbers in some cases, and unsigned numbers in other cases, and it is also possible that one is a signed number and the other is an unsigned number; In some occasions, not only multiply-accumulate operations are required, but also multiplication operations of operands with different bit widths, multiply-accumulate operations, and parallel multiply-accumulate operations are required. At...

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): G06F7/544
Inventor 刘鹏夏冰洁姚庆栋
Owner ZHEJIANG 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
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