You copied the Doc URL to your clipboard.

NEON intrinsics for absolute difference

These intrinsics provide operations including absolute difference.

Absolute difference between the arguments: vabd{q}_<type>. Vr[i] = | Va[i] - Vb[i] |

int8x8_t    vabd_s8(int8x8_t a, int8x8_t b);         // VABD.S8 d0,d0,d0 
int16x4_t   vabd_s16(int16x4_t a, int16x4_t b);      // VABD.S16 d0,d0,d0
int32x2_t   vabd_s32(int32x2_t a, int32x2_t b);      // VABD.S32 d0,d0,d0
uint8x8_t   vabd_u8(uint8x8_t a, uint8x8_t b);       // VABD.U8 d0,d0,d0 
uint16x4_t  vabd_u16(uint16x4_t a, uint16x4_t b);    // VABD.U16 d0,d0,d0
uint32x2_t  vabd_u32(uint32x2_t a, uint32x2_t b);    // VABD.U32 d0,d0,d0
float32x2_t vabd_f32(float32x2_t a, float32x2_t b);  // VABD.F32 d0,d0,d0
int8x16_t   vabdq_s8(int8x16_t a, int8x16_t b);      // VABD.S8 q0,q0,q0 
int16x8_t   vabdq_s16(int16x8_t a, int16x8_t b);     // VABD.S16 q0,q0,q0
int32x4_t   vabdq_s32(int32x4_t a, int32x4_t b);     // VABD.S32 q0,q0,q0
uint8x16_t  vabdq_u8(uint8x16_t a, uint8x16_t b);    // VABD.U8 q0,q0,q0 
uint16x8_t  vabdq_u16(uint16x8_t a, uint16x8_t b);   // VABD.U16 q0,q0,q0
uint32x4_t  vabdq_u32(uint32x4_t a, uint32x4_t b);   // VABD.U32 q0,q0,q0
float32x4_t vabdq_f32(float32x4_t a, float32x4_t b); // VABD.F32 q0,q0,q0

Absolute difference - long

int16x8_t  vabdl_s8(int8x8_t a, int8x8_t b);      // VABDL.S8 q0,d0,d0 
int32x4_t  vabdl_s16(int16x4_t a, int16x4_t b);   // VABDL.S16 q0,d0,d0
int64x2_t  vabdl_s32(int32x2_t a, int32x2_t b);   // VABDL.S32 q0,d0,d0
uint16x8_t vabdl_u8(uint8x8_t a, uint8x8_t b);    // VABDL.U8 q0,d0,d0 
uint32x4_t vabdl_u16(uint16x4_t a, uint16x4_t b); // VABDL.U16 q0,d0,d0
uint64x2_t vabdl_u32(uint32x2_t a, uint32x2_t b); // VABDL.U32 q0,d0,d0

Absolute difference and accumulate: vaba{q}_<type>. Vr[i] = Va[i] + | Vb[i] - Vc[i] |

int8x8_t   vaba_s8(int8x8_t a, int8x8_t b, int8x8_t c);         // VABA.S8 d0,d0,d0 
int16x4_t  vaba_s16(int16x4_t a, int16x4_t b, int16x4_t c);     // VABA.S16 d0,d0,d0
int32x2_t  vaba_s32(int32x2_t a, int32x2_t b, int32x2_t c);     // VABA.S32 d0,d0,d0
uint8x8_t  vaba_u8(uint8x8_t a, uint8x8_t b, uint8x8_t c);      // VABA.U8 d0,d0,d0 
uint16x4_t vaba_u16(uint16x4_t a, uint16x4_t b, uint16x4_t c);  // VABA.U16 d0,d0,d0
uint32x2_t vaba_u32(uint32x2_t a, uint32x2_t b, uint32x2_t c);  // VABA.U32 d0,d0,d0
int8x16_t  vabaq_s8(int8x16_t a, int8x16_t b, int8x16_t c);     // VABA.S8 q0,q0,q0 
int16x8_t  vabaq_s16(int16x8_t a, int16x8_t b, int16x8_t c);    // VABA.S16 q0,q0,q0
int32x4_t  vabaq_s32(int32x4_t a, int32x4_t b, int32x4_t c);    // VABA.S32 q0,q0,q0
uint8x16_t vabaq_u8(uint8x16_t a, uint8x16_t b, uint8x16_t c);  // VABA.U8 q0,q0,q0 
uint16x8_t vabaq_u16(uint16x8_t a, uint16x8_t b, uint16x8_t c); // VABA.U16 q0,q0,q0
uint32x4_t vabaq_u32(uint32x4_t a, uint32x4_t b, uint32x4_t c); // VABA.U32 q0,q0,q0

Absolute difference and accumulate - long

int16x8_t  vabal_s8(int16x8_t a, int8x8_t b, int8x8_t c);       // VABAL.S8 q0,d0,d0 
int32x4_t  vabal_s16(int32x4_t a, int16x4_t b, int16x4_t c);    // VABAL.S16 q0,d0,d0
int64x2_t  vabal_s32(int64x2_t a, int32x2_t b, int32x2_t c);    // VABAL.S32 q0,d0,d0
uint16x8_t vabal_u8(uint16x8_t a, uint8x8_t b, uint8x8_t c);    // VABAL.U8 q0,d0,d0 
uint32x4_t vabal_u16(uint32x4_t a, uint16x4_t b, uint16x4_t c); // VABAL.U16 q0,d0,d0
uint64x2_t vabal_u32(uint64x2_t a, uint32x2_t b, uint32x2_t c); // VABAL.U32 q0,d0,d0