RU95104047A - Associative multiplier - Google Patents

Associative multiplier

Info

Publication number
RU95104047A
RU95104047A RU95104047/09A RU95104047A RU95104047A RU 95104047 A RU95104047 A RU 95104047A RU 95104047/09 A RU95104047/09 A RU 95104047/09A RU 95104047 A RU95104047 A RU 95104047A RU 95104047 A RU95104047 A RU 95104047A
Authority
RU
Russia
Prior art keywords
diagonal
ones
sums
units
zeros
Prior art date
Application number
RU95104047/09A
Other languages
Russian (ru)
Inventor
Г.И. Васильев
М.В. Самуйлов
Original Assignee
Серпуховское высшее военное командно-инженерное училище ракетных войск имени Ленинского комсомола
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Серпуховское высшее военное командно-инженерное училище ракетных войск имени Ленинского комсомола filed Critical Серпуховское высшее военное командно-инженерное училище ракетных войск имени Ленинского комсомола
Priority to RU95104047/09A priority Critical patent/RU95104047A/en
Publication of RU95104047A publication Critical patent/RU95104047A/en

Links

Landscapes

  • Complex Calculations (AREA)

Abstract

FIELD: automation and computer engineering. SUBSTANCE: device provides simultaneous processing of all bits of two factors so that during one clock cycle, presence of odd number of ones in diagonal section results in generation of all one values for partial sums in each diagonal. Carries are generated using position of border between zeros and ones in ordered diagonal section. Final calculation of product is achieved during subsequent additional cycles, which number is determined by absence of ones between carries. Corresponding device has factors registers 1, 2, unit of registers or associative memory unit 3, group of units 5 which process diagonal sections, unit 10 which stores values of intermediate sums and carries, and synchronization unit 14. Each unit 5 orders diagonal section so that ones go before zeros. Analysis of diagonal sections for odd number of ones and position of border between zeros and ones provides partial sums and carries when partial sums are added. Final product is calculated during one main and several additional cycles. EFFECT: increased speed.

Claims (1)

Цель изобретения - сокращение временных затрат на получение произведения двух сомножителей. Это достигается путем одновременной обработки всех разрядов двух сомножителей так, что в один такт работы по признаку нечетности числа единиц в диагональном срезе формируются единичные значения частичных сумм в каждой диагонали, а по признаку положения границы между нулями и единицами в упорядоченном диагональном срезе формируются переносы. Окончательное формирование произведения происходит на последующих дополнительных тактах работы, число которых определяется отсутствием единиц среди переносов. Указанный способ реализуется использованием устройства, включающего регистры сомножителей 1, 2, блок регистров или блок ассоциативной памяти 3, группу блоков обработки диагональных срезов 5, блок хранения значений промежуточных сумм и переносов 10, а также блок синхронизации 14. В каждом блоке 5 диагональные срезы частичных сумм упорядочиваются так, что вначале располагаются подряд все единицы, а затем все нули. Затем по результатам анализа диагональных срезов на нечетность числа единиц и на положение границы между "0"" и "1" формируют суммы и переносы при сложении частичных сумм. За один основной и несколько дополнительных тактов работы устройства получают окончательное произведение.The purpose of the invention is to reduce the time required to obtain the product of two factors. This is achieved by simultaneously processing all the bits of two factors so that in one cycle of work, on the basis of the odd number of units in the diagonal slice, unit values of the partial sums in each diagonal are formed, and transfers are formed on the basis of the position of the boundary between zeros and ones in the ordered diagonal slice. The final formation of the work takes place at subsequent additional steps of the work, the number of which is determined by the absence of units among transfers. The indicated method is implemented using a device including registers of factors 1, 2, a register block or an associative memory block 3, a group of processing units for diagonal slices 5, a block for storing values of intermediate sums and transfers 10, and also a synchronization block 14. In each block 5, diagonal partial slices sums are ordered so that first all units are arranged in a row, and then all zeros. Then, according to the results of the analysis of diagonal slices for the odd number of units and the position of the border between “0” and “1”, sums and transfers are formed when partial sums are added. For one main and several additional clock cycles, the devices receive the final product.
RU95104047/09A 1995-03-21 1995-03-21 Associative multiplier RU95104047A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
RU95104047/09A RU95104047A (en) 1995-03-21 1995-03-21 Associative multiplier

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
RU95104047/09A RU95104047A (en) 1995-03-21 1995-03-21 Associative multiplier

Publications (1)

Publication Number Publication Date
RU95104047A true RU95104047A (en) 1997-01-20

Family

ID=48433075

Family Applications (1)

Application Number Title Priority Date Filing Date
RU95104047/09A RU95104047A (en) 1995-03-21 1995-03-21 Associative multiplier

Country Status (1)

Country Link
RU (1) RU95104047A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
RU2525477C2 (en) * 2010-12-09 2014-08-20 Антон Николаевич Кривоногов Method of multiplying decimal numbers

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
RU2525477C2 (en) * 2010-12-09 2014-08-20 Антон Николаевич Кривоногов Method of multiplying decimal numbers

Similar Documents

Publication Publication Date Title
ES8302333A1 (en) Data processing apparatus.
EP0675450A3 (en) Array processor for intensive numeric computation
KR960042330A (en) Fourier transform arithmetic unit and method
RU95104047A (en) Associative multiplier
JP3323312B2 (en) Accelerated test pattern generator
US4546445A (en) Systolic computational array
Chaitin An improvement on a theorem of EF Moore
JPS58129653A (en) Multiplication system
JPS62166471A (en) Parallel processing system for image data
JPH0214363A (en) Fast fourier transform method and arithmetic apparatus therefor
US4631662A (en) Scanning alarm electronic processor
KR970025177A (en) Method and apparatus for performing DCT / DST / DHT using unified systolic array structure
SU1277095A1 (en) Device for taking sum of n-bit binary numbers
SU976442A1 (en) Device for scheduling tasks for processors
SU866561A1 (en) Device for quick fourier transform
Valero-Garcia et al. Systematic hardware adaptation of systolic algorithms
SU736112A1 (en) Fourier coefficients computing arrangement
SU564638A1 (en) Device for solving linear algebraic equations systems
JPH07271767A (en) Vector arithmetic processor
RU2047217C1 (en) Device implementing modulo-three convolution
Sundaram et al. T-algorithm-based logic simulation on distributed systems.
SU1606973A1 (en) Device for sorting numbers
RU1829119C (en) Device to count number of units in binary-decimal code system
JPS63170788A (en) Image processor
JPS61246837A (en) Parallel multiplier