Integer shuffles — byte, word, dword (per 128-bit lane, mask or imm8 controlled)

Float shuffles — single / double precision, two source registers, imm8 controlled

128-bit lane shuffles — pick whole 128-bit lanes from two sources, imm8 controlled, k-mask supported

Broadcasts — replicate one element or a 128-/256-bit unit across the destination, k-mask supported

Sign / zero extension — widen each packed byte/word/dword element to a larger element, k-mask supported

Unpack low — interleave the low halves of two sources per 128-bit lane, k-mask supported

Unpack high — interleave the high halves of two sources per 128-bit lane, k-mask supported

Packed equality — all-ones or all-zeros per element, or an opmask in the EVEX form

Byte sign mask — pack the top bit of every byte into a general-purpose register

Insert — copy one byte/dword/qword from a general-purpose register or memory into the XMM slot picked by imm8

String compare — the imm8-controlled comparison matrix, its aggregation into IntRes1/IntRes2, and the index it puts in ECX

Ternary logic — any 3-input Boolean function, bit by bit, with imm8 as its truth table; k-mask and broadcast supported

Predicate compares into an opmask — imm8 picks EQ/LT/LE/FALSE/NEQ/NLT/NLE/TRUE, signed or unsigned

Opmask shifts — move the mask bits by imm8, shifting zeros in and dropping what falls off

Opmask concatenation — glue two masks into one twice as wide; no interleaving despite the name