|Important||This document may not represent best practices for current development, links to downloads and other resources may no longer be valid. Current recommended version can be found here. ArchiveDisclaimer|
Guarantees that every preceding store is globally visible before any subsequent store.
On the x64 platform, this routine generates an instruction that is a faster store fence than the sfence instruction. Use this intrinsic instead of _mm_sfence on the x64 platform.
This routine is only available as an intrinsic.