SU1672444A1 - Optical module in the residue classes system - Google Patents

Abstract

The invention relates to computing and can be used in the manufacture of digital optical processors using residual arithmetic. The purpose of the invention is to expand the functionality by allowing the conversion of the electric code of the operand to the optical code and the calculation of the correction of the rank of the operand on this basis. The module contains a matrix of optical switches (OK) 1 and photodetectors (F) 2 (M lines and (M - 1) OK and F in each row), the inputs of the first 3 and second 4 operands, the optical code outputs of the operand 5 and the outputs of the rank correction operand 6. Both operands come in a unitary code, the first operand is in the optical code, the second is in the electrical code, and the optical code of the second operand is generated at outputs 5, and the correction code of the rank ΔR _and outputs of the first operand at A = ( α ₁ , α ₂ , ... α _m , ... α _N , R _a ) → (α ₁ , α ₂ , ... β _m , ... α _N , R _a + ΔR _a ). 1 il.

Description

The invention relates to computer technology and can be used in the manufacture of digital optical processors using arithmetic residues in the pulse-position representation of the operands.

The purpose of the invention is the expansion of functionality by providing the ability to convert the electrical code of the operand into an optical code and calculate the correction of the rank of the operand on this basis.

The drawing shows a diagram of the module.

The module contains a matrix of M rows of optical switches 1 and photodetectors 2 along (M-1) switches and photodetectors in each row, where M is the base of the RNS, inputs of the first 3 and second 4 operands, outputs 5 of the optical code of the operand and outputs 6 of the correction of the operand rank . Thin lines indicate electrical connections, and thickened lines indicate optical ones.

The first operand Ohm arrives at inputs 3 in a unitary (one of M) optical code, the second operand _{β Μ} arrives at inputs 4 in a unitary electric code. If a _m * β _Μ , then the Ohm optical switch of the βπth row switches and the optical signal from the input of the 3 ^ module goes to the “mth photodetector of 2 β ^ -Λ lines, which generates an electrical signal at the corresponding output of the 6th rank correction . At the same time, an optical signal appears on the β _Μ output 5 of the optical code of the operand.

If "m - β". then the optical pulse does not experience permutations at outputs 5, and no electrical signal is generated at outputs 6.

Thus, the module generates the value β _Μ in the optical code and calculates the correction for the rank Δγα of the first operand when converting A = (od. Od ..... "m ....." p, gd) - "(od, od, .... / 3 _m ..... "p. Ha + Δγα) ..

A set of η described modules and the corresponding number of adders can be used to compare numbers

A = ("1. OD ....." p, GA)

And B = (/? 1, / ¾ ..... βη, GV).

where η is the base of RNS;

α, βι - residues on the bases of RNS:

ha, gv - ranks of numbers A and B.

Moreover, the number A is reduced to the form using the described modules

А = (/31./¾...../?пг'а).

where rU is the new rank of the number A, which is calculated as r! * = rd + 2 (/ 3ι-αι) -§-, 1 = 1 ^g where B, is the orthogonal basis of the основанияth base:

P is the range of RNS;

and B> if g’a <gv.

A = B, 'if r'a = gv,

A <B, if r * a> gv.

The set of modules can also be used to convert the number A from JUICE to a positional number system:

A ^ (0, 0 ..... 0. Ha).

then A = r a -P.

Claims (1)

Claim An optical module in a system of residual classes, containing Μ (M-1) photodetectors and a matrix of M rows of optical switches, where M is the base of the system of residual classes, and each row contains (M-1) optical switches, an input of the Ιth discharge of the first operand module, where 1 = 1, 2 ..... (M-1), is connected to the optical information input of the optical switch i-ro of the discharge of the first row of the matrix, the first output of the optical switch j-ro of the discharge of the Kth row of the matrix, where J = 0, 1 ..... (M-1), K = 0. 1 ..... (M-2), j * K. J * K-1, connected to the optical information input o cally switch i-ro discharge (K + 1) -th row, j-ro input of the second module operand bit is connected to electrical control inputs of all optical switches j-th row of the matrix, wherein. that, in order to expand the functionality by providing the possibility of converting the operand’s electrical code into an optical code and calculating the correction of the operand’s rank on this basis, the zero-input of the first operand of the module is connected to the information optical input of the optical switch of the zero discharge of the first row of the matrix, the first output optical switch (i-1) -ro discharge (M-1) -th row of the matrix is the output of the (i-1) -ro discharge optical code of the second operand of the module, the first output of the optical the mutator of the (M-1) -th discharge of the (M-2) -th row of the matrix is the output of the (M-1) -th discharge of the optical code of the second operand of the module, the first output of the optical switch of the Ι-th discharge of the (1-1) th row matrices, where I = 1, 2 ..... (M-2), is connected to the optical information input of the optical switch of the 1st category of the (1 + 1) -th row of the matrix, the second output of the optical switch of the q-ro discharge is j- nth row of the matrix, where q = 0, 1 ..... (M-1), q] is connected to the input of the q-ro photodetector of the jth row of the matrix, the second outputs of all optical switches of the Kth row of the matrix are connected to the input optic switch of the Kth category (K + 1) -th row of the matrix, the second outputs of all optical 5 switches (M-1) -th row of the matrix are connected to the output of the (M-1) -th category of the optical code of the second operand of the module, the output of the correction Δ module rank, where Δ = ± 1, ± 2 ..... ± (M - 1), is connected to 10 outputs of such S-x photodetectors of the jth row of the matrix for which S = j-A