SDOT (4-way, multiple and indexed vector)

Multi-vector signed integer dot-product by indexed element

The signed integer dot product instruction computes the dot product of four signed 8-bit or 16-bit integer values held in each 32-bit or 64-bit element of the two or four first source vectors and four signed 8-bit or 16-bit integer values in the corresponding indexed 32-bit or 64-bit element of the second source vector. The widened dot product result is destructively added to the corresponding 32-bit or 64-bit element of the ZA single-vector groups.

The groups within the second source vector are specified using an immediate element index which selects the same group position within each 128-bit vector segment. The index range is from 0 to one less than the number of groups per 128-bit segment, encoded in 1 to 2 bits depending on the size of the group. The vector numbers forming the single-vector group within each half of or each quarter of the ZA array are selected by the sum of the vector select register and immediate offset, modulo half or quarter the number of ZA array vectors.

The vector group symbol, VGx2 or VGx4, indicates that the ZA operand consists of two or four ZA single-vector groups respectively. The vector group symbol is preferred for disassembly, but optional in assembler source code.

This instruction is unpredicated.

ID_AA64SMFR0_EL1.I16I64 indicates whether the 16-bit integer variant is implemented.

It has encodings from 4 classes: Two ZA single-vectors of 32-bit elements , Two ZA single-vectors of 64-bit elements , Four ZA single-vectors of 32-bit elements and Four ZA single-vectors of 64-bit elements

Two ZA single-vectors of 32-bit elements
(FEAT_SME2)

313029282726252423222120191817161514131211109876543210
110000010101Zm0Rv1i2Zn100off3
U

SDOT ZA.S[<Wv>, <offs>{, VGx2}], { <Zn1>.B-<Zn2>.B }, <Zm>.B[<index>]

if !HaveSME2() then UNDEFINED; integer v = UInt('010':Rv); constant integer esize = 32; integer n = UInt(Zn:'0'); integer m = UInt('0':Zm); integer offset = UInt(off3); integer index = UInt(i2); constant integer nreg = 2;

Two ZA single-vectors of 64-bit elements
(FEAT_SME_I16I64)

313029282726252423222120191817161514131211109876543210
110000011101Zm0Rv00i1Zn001off3
U

SDOT ZA.D[<Wv>, <offs>{, VGx2}], { <Zn1>.H-<Zn2>.H }, <Zm>.H[<index>]

if !(HaveSME2() && HaveSMEI16I64()) then UNDEFINED; integer v = UInt('010':Rv); constant integer esize = 64; integer n = UInt(Zn:'0'); integer m = UInt('0':Zm); integer offset = UInt(off3); integer index = UInt(i1); constant integer nreg = 2;

Four ZA single-vectors of 32-bit elements
(FEAT_SME2)

313029282726252423222120191817161514131211109876543210
110000010101Zm1Rv1i2Zn0100off3
U

SDOT ZA.S[<Wv>, <offs>{, VGx4}], { <Zn1>.B-<Zn4>.B }, <Zm>.B[<index>]

if !HaveSME2() then UNDEFINED; integer v = UInt('010':Rv); constant integer esize = 32; integer n = UInt(Zn:'00'); integer m = UInt('0':Zm); integer offset = UInt(off3); integer index = UInt(i2); constant integer nreg = 4;

Four ZA single-vectors of 64-bit elements
(FEAT_SME_I16I64)

313029282726252423222120191817161514131211109876543210
110000011101Zm1Rv00i1Zn0001off3
U

SDOT ZA.D[<Wv>, <offs>{, VGx4}], { <Zn1>.H-<Zn4>.H }, <Zm>.H[<index>]

if !(HaveSME2() && HaveSMEI16I64()) then UNDEFINED; integer v = UInt('010':Rv); constant integer esize = 64; integer n = UInt(Zn:'00'); integer m = UInt('0':Zm); integer offset = UInt(off3); integer index = UInt(i1); constant integer nreg = 4;

Assembler Symbols

<Wv>

Is the 32-bit name of the vector select register W8-W11, encoded in the "Rv" field.

<offs>

Is the vector select offset, in the range 0 to 7, encoded in the "off3" field.

<Zn1>

For the two ZA single-vectors of 32-bit elements and two ZA single-vectors of 64-bit elements variant: is the name of the first scalable vector register of the first source multi-vector group, encoded as "Zn" times 2.

For the four ZA single-vectors of 32-bit elements and four ZA single-vectors of 64-bit elements variant: is the name of the first scalable vector register of the first source multi-vector group, encoded as "Zn" times 4.

<Zn4>

Is the name of the fourth scalable vector register of the first source multi-vector group, encoded as "Zn" times 4 plus 3.

<Zn2>

Is the name of the second scalable vector register of the first source multi-vector group, encoded as "Zn" times 2 plus 1.

<Zm>

Is the name of the second source scalable vector register Z0-Z15, encoded in the "Zm" field.

<index>

For the four ZA single-vectors of 32-bit elements and two ZA single-vectors of 32-bit elements variant: is the immediate index of a 32-bit group of four 8-bit values within each 128-bit vector segment, in the range 0 to 3, encoded in the "i2" field.

For the four ZA single-vectors of 64-bit elements and two ZA single-vectors of 64-bit elements variant: is the immediate index of a 64-bit group of four 16-bit values within each 128-bit vector segment, in the range 0 to 1, encoded in the "i1" field.

Operation

CheckStreamingSVEAndZAEnabled(); constant integer VL = CurrentVL; constant integer elements = VL DIV esize; integer vectors = VL DIV 8; integer vstride = vectors DIV nreg; integer eltspersegment = 128 DIV esize; bits(32) vbase = X[v, 32]; integer vec = (UInt(vbase) + offset) MOD vstride; bits(VL) result; for r = 0 to nreg-1 bits(VL) operand1 = Z[n+r, VL]; bits(VL) operand2 = Z[m, VL]; bits(VL) operand3 = ZAvector[vec, VL]; for e = 0 to elements-1 bits(esize) sum = Elem[operand3, e, esize]; integer segmentbase = e - (e MOD eltspersegment); integer s = segmentbase + index; for i = 0 to 3 integer element1 = SInt(Elem[operand1, 4 * e + i, esize DIV 4]); integer element2 = SInt(Elem[operand2, 4 * s + i, esize DIV 4]); sum = sum + element1 * element2; Elem[result, e, esize] = sum; ZAvector[vec, VL] = result; vec = vec + vstride;


Internal version only: aarchmrs v2023-12_rel, pseudocode v2023-12_rel, sve v2023-12_rel ; Build timestamp: 2023-12-15T16:46

Copyright © 2010-2023 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.