BFCLAMP

BFloat16 clamp to minimum/maximum number

This instruction clamps each BFloat16 element in the destination vector to between the BFloat16 minimum value in the corresponding element of the first source vector and the BFloat16 maximum value in the corresponding element of the second source vector and destructively places the clamped results in the corresponding elements of the destination vector.

Regardless of the value of FPCR.AH, the behavior is as follows for each minimum number and maximum number operation:

This instruction follows SVE2 non-widening BFloat16 numerical behaviors.

This instruction is unpredicated.

ID_AA64ZFR0_EL1.B16B16 indicates whether this instruction is implemented.

SVE2
(FEAT_SVE_B16B16)

313029282726252423222120191817161514131211109876543210
01100100001Zm001001ZnZd
size

Encoding

BFCLAMP <Zd>.H, <Zn>.H, <Zm>.H

Decode for this encoding

if !IsFeatureImplemented(FEAT_SVE_B16B16) then EndOfDecode(Decode_UNDEF); end; let n : integer = UInt(Zn); let m : integer = UInt(Zm); let d : integer = UInt(Zd);

Assembler Symbols

<Zd>

Is the name of the destination scalable vector register, encoded in the "Zd" 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, encoded in the "Zm" field.

Operation

if IsFeatureImplemented(FEAT_SME2) then CheckSVEEnabled(); else CheckNonStreamingSVEEnabled(); end; let VL : integer{} = CurrentVL(); let elements : integer = VL DIV 16; var result : bits(VL); let operand1 : bits(VL) = Z{}(n); let operand2 : bits(VL) = Z{}(m); let operand3 : bits(VL) = Z{}(d); for e = 0 to elements-1 do let element1 : bits(16) = operand1[e*:16]; let element2 : bits(16) = operand2[e*:16]; let element3 : bits(16) = operand3[e*:16]; let maxelement : bits(16) = BFMaxNum{}(element1, element3, FPCR()); result[e*:16] = BFMinNum{16}(maxelement, element2, FPCR()); end; Z{VL}(d) = 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:


2026-03_rel 2026-03-26 20:48:11

Copyright © 2010-2026 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.