Instruction category | Instructions | Benefits |
Packed DWORD Multiplies | PMULLD, PMULDQ | Improved automated compiler vectorization |
Floating Point Dot Product | DPPS, DPPD | 3D content creation, gaming, support for languages such as CG and HLSL |
Multi-packed sum of absolute diffs& min pos | MPSADBW, PHMINPOSUW | Video processing |
Packed Blending | BLENDPS, BLENDPD, BLENDVPS, BLENDVPD, PBLENDVB, PBLENDDW | Compiler vectorization and applications such as video processing, multi-media and gaming |
Packed Integer Min and Max | PMINSB, PMAXSB, PMINUW, PMAXUW, PMINUD, PMAXUD, PMINDS, PMAXSD | Compiler vectorization and applications such as video processing, multi-media and gaming |
Floating Point Round | ROUNDPS, ROUNDSS, ROUNDPD, ROUNDSD | Image processing, graphics, video processing, 2D/3D applications, multimedia, and gaming. |
Register Register Insertion/ExtractionInsertion/Extraction | INSERTPS, PINSRB, PINSRD, PINSRQ, EXTRACTPS, PEXTRB, PEXTRD, PEXTRW, PEXTRQ | Compiler vectorization and applications such as video processing, multi-media and gaming |
Packed Format Conversion | PMOVSXBW, PMOVZXBW, PMOVSXBD, PMOVZXBD, PMOVSXBQ, PMOVZXBQ, PMOVSXWD, PMOVZXWD, PMOVSXWQ, PMOVZXWQ, PMOVSXDQ, PMOVZXDQ | Compiler vectorization and applications such as video processing, multi-media and gaming |
Streaming load | MOVNTDQA | Video processing, imaging, data sharing apps with GPU |
Packed Test & Set | PTESTPTEST | Compiler vectorization and applications such as video processing, multi-media and gaming |
Packed Compare for Equal | PCMPEQQ | Compiler vectorization and applications such as video processing, multi-media and gaming |
Pack DWORD to Unsigned WORD | PACKUSDW | Compiler vectorization and applications such as video processing, multi-media and gaming |
Intel Penryn SSE4 instruction set
Posted on Wednesday, April 18 2007 @ 14:18 CEST by Thomas De Maesschalck