Справочник Пользователя для AMD 250
Appendix E
SSE and SSE2 Optimizations
363
Software Optimization Guide for AMD64 Processors
25112
Rev. 3.06
September 2005
E.7
Explicit Load Instructions
Optimization
Use
movlpd xmm1, mem64
when loading a scalar FPD value from memory.
Application
This optimization applies to:
•
32-bit software
•
64-bit software
Rationale
The
movlpd xmm1, mem64
instruction is more efficient than
movsd xmm1, mem64
. Use MOVSD only
if you need to ensure that the upper half of XMM1 is also set to FPD format, perhaps because a vector
operation is planned on the register.
operation is planned on the register.
When loading a scalar FPS value from memory, use MOVSS.