Developer Guide and Reference

Contents

_mm256_cvtepi8_epi16/32/64

Performs packed move with sign-extend on 8-bit signed integers to 16/32/64-bit integers. The corresponding Intel® AVX2 instruction is
VPMOVSXBW
,
VPMOVSXBD
, or
VPMOVSXBQ
.

Syntax

extern __m256i _mm256_cvtepi8_epi16(__m128i s1);
extern __m256i _mm256_cvtepi8_epi32(__m128i s1);
extern __m256i _mm256_cvtepi8_epi64(__m128i s1);
Arguments
s1
128-bit integer source vector used for the operation
Description
Performs a packed move with sign-extend operation to convert 8-bit [byte] integers in the low bytes of the source vector,
s1
, to 16-bit [word], 32-bit [doubleword], or 64-bit [quadword] integers, which are stored as packed signed word/doubleword/quadword integers in the destination vector.
Returns
Result of the sign-extend operation.