Available on x86 and target feature
sse only.Expand description
Stores four 32-bit floats into memory. There are no restrictions on memory
alignment. For aligned memory _mm_store_ps may be
faster.
This corresponds to instructions VMOVUPS / MOVUPS.