FMLALT (indexed, FP8 to FP16)

8-bit floating-point multiply-add long to half-precision (top, indexed)

This 8-bit floating-point multiply-add long instruction widens the odd 8-bit elements in the first source vector and the indexed element from the corresponding 128-bit segment in the second source vector to half-precision format and multiplies the corresponding elements. The intermediate products are scaled by 2-UInt(FPMR.LSCALE[3:0]) before being destructively added without intermediate rounding to the half-precision elements of the destination vector that overlap with the corresponding 8-bit floating-point elements in the first source vector. The 8-bit floating-point encoding format for the elements of the first source vector and the second source vector is selected by FPMR.F8S1 and FPMR.F8S2 respectively.

This instruction is unpredicated.

SVE2
(FEAT_FP8FMA)

313029282726252423222120191817161514131211109876543210
01100100101i4hZm0101i4lZnZda
T

FMLALT <Zda>.H, <Zn>.B, <Zm>.B[<imm>]

if !HaveSVE2FP8FMA() then UNDEFINED; constant integer n = UInt(Zn); constant integer m = UInt(Zm); constant integer da = UInt(Zda); constant integer index = UInt(i4h:i4l);

Assembler Symbols

<Zda>

Is the name of the third source and destination scalable vector register, encoded in the "Zda" field.

<Zn>

Is the name of the first source scalable vector register, encoded in the "Zn" field.

<Zm>

Is the name of the second source scalable vector register Z0-Z7, encoded in the "Zm" field.

<imm>

Is the immediate index, in the range 0 to 15, encoded in the "i4h:i4l" fields.

Operation

CheckFPMREnabled(); if IsFeatureImplemented(FEAT_FP8FMA) then CheckSVEEnabled(); else CheckStreamingSVEEnabled(); constant integer VL = CurrentVL; constant integer elements = VL DIV 16; constant integer eltspersegment = 128 DIV 16; constant bits(VL) operand1 = Z[n, VL]; constant bits(VL) operand2 = Z[m, VL]; constant bits(VL) operand3 = Z[da, VL]; bits(VL) result; for e = 0 to elements-1 constant integer segmentbase = e - (e MOD eltspersegment); constant integer s = 2 * segmentbase + index; constant bits(8) element1 = Elem[operand1, 2 * e + 1, 8]; constant bits(8) element2 = Elem[operand2, s, 8]; constant bits(16) element3 = Elem[operand3, e, 16]; Elem[result, e, 16] = FP8MulAddFP(element3, element1, element2, FPCR, FPMR); Z[da, VL] = result;

Operational information

This instruction might be immediately preceded in program order by a MOVPRFX instruction. The MOVPRFX must conform to all of the following requirements, otherwise the behavior of the MOVPRFX and this instruction is constrained unpredictable:


Internal version only: aarchmrs v2024-03_relA, pseudocode v2024-03_rel, sve v2024-03_rel ; Build timestamp: 2024-03-26T09:45

Copyright © 2010-2024 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.