movntps
Move Non-Temporal Packed Single
MOVNTPS m128, xmm
Stores float vectors directly to RAM, bypassing cache.
Details
The Move Non-Temporal Packed Single instruction stores float vectors directly to RAM, bypassing cache.
Pseudocode Operation
// Stores float vectors directly to RAM, bypassing cache
Example
MOVNTPS [rbp-16], xmm0
Encoding
Binary Layout
0F
+0
2B
+1
Operands
-
dest
128-bit memory operand -
src
128-bit SSE/AVX register (XMM)