NASM 2.05 based x86 Instruction Reference[ch028]
A.5.9 ANDNPD: Bitwise Logical AND NOT of Packed Double-Precision FP Values ANDNPD xmm1,xmm2/mem128 ; 66 0F 55 /r [WILLAMETTE,SSE2] ANDNPD inverts the bits of the two double-precision floating-point values in the destination register, and then performs a logical AND between the two double-precision floating-point values in the source operand and the temporary inverted result, storing the result in the destination register. dst[0-63] := src[0-63] AND NOT dst[0-63], dst[64-127] := src[64-127] AND NOT dst[64-127]. The destination is an XMM register. The source operand can be either an XMM register or a 128-bit memory location.