AI Engine-ML Intrinsics User Guide (v2025.1)
Loading...
Searching...
No Matches
Emulated Multiply-accumulate of 16-bit Complex Brain Floating-Point

Elementwise matrix multiplications emulated on top of bfloat16. More...

Overview

Elementwise matrix multiplications emulated on top of bfloat16.

Multiplication of (2x8) with (8x2) bfloat16 and complex bfloat16

v4caccfloat mul_2x8_8x2 (v16bfloat16 a, v16cbfloat16 b)
 
v4caccfloat negmul_2x8_8x2 (v16bfloat16 a, v16cbfloat16 b)
 
v4caccfloat mac_2x8_8x2 (v16bfloat16 a, v16cbfloat16 b, v4caccfloat acc1)
 
v4caccfloat msc_2x8_8x2 (v16bfloat16 a, v16cbfloat16 b, v4caccfloat acc1)
 
v4caccfloat addmac_2x8_8x2 (v16bfloat16 a, v16cbfloat16 b, v4caccfloat acc1, v4caccfloat acc2)
 
v4caccfloat addmsc_2x8_8x2 (v16bfloat16 a, v16cbfloat16 b, v4caccfloat acc1, v4caccfloat acc2)
 

Multiplication of (2x8) with (8x2) complex bfloat16 and bfloat16

v4caccfloat mul_2x8_8x2 (v16cbfloat16 a, v16bfloat16 b)
 
v4caccfloat negmul_2x8_8x2 (v16cbfloat16 a, v16bfloat16 b)
 
v4caccfloat mac_2x8_8x2 (v16cbfloat16 a, v16bfloat16 b, v4caccfloat acc1)
 
v4caccfloat msc_2x8_8x2 (v16cbfloat16 a, v16bfloat16 b, v4caccfloat acc1)
 
v4caccfloat addmac_2x8_8x2 (v16cbfloat16 a, v16bfloat16 b, v4caccfloat acc1, v4caccfloat acc2)
 
v4caccfloat addmsc_2x8_8x2 (v16cbfloat16 a, v16bfloat16 b, v4caccfloat acc1, v4caccfloat acc2)
 

Multiplication of (2x8) with (8x2) complex bfloat16 and complex bfloat16

v4caccfloat mul_2x8_8x2 (v16cbfloat16 a, v16cbfloat16 b)
 
v4caccfloat negmul_2x8_8x2 (v16cbfloat16 a, v16cbfloat16 b)
 
v4caccfloat mac_2x8_8x2 (v16cbfloat16 a, v16cbfloat16 b, v4caccfloat acc1)
 
v4caccfloat msc_2x8_8x2 (v16cbfloat16 a, v16cbfloat16 b, v4caccfloat acc1)
 
v4caccfloat addmac_2x8_8x2 (v16cbfloat16 a, v16cbfloat16 b, v4caccfloat acc1, v4caccfloat acc2)
 
v4caccfloat addmsc_2x8_8x2 (v16cbfloat16 a, v16cbfloat16 b, v4caccfloat acc1, v4caccfloat acc2)
 

Multiplication of (2x8) with (8x2) bfloat16 and complex bfloat16 with dynamic negation of multiplication result

v4caccfloat mul_2x8_8x2_conf (v16bfloat16 a, v16cbfloat16 b, int sub_mul)
 
v4caccfloat negmul_2x8_8x2_conf (v16bfloat16 a, v16cbfloat16 b, int sub_mul)
 

Multiplication of (2x8) with (8x2) complex bfloat16 and bfloat16 with dynamic negation of multiplication result

v4caccfloat mul_2x8_8x2_conf (v16cbfloat16 a, v16bfloat16 b, int sub_mul)
 
v4caccfloat negmul_2x8_8x2_conf (v16cbfloat16 a, v16bfloat16 b, int sub_mul)
 

Multiplication of (2x8) with (8x2) complex bfloat16 and complex bfloat16 with dynamic negation of multiplication result

v4caccfloat mul_2x8_8x2_conf (v16cbfloat16 a, v16cbfloat16 b, int sub_mul)
 
v4caccfloat negmul_2x8_8x2_conf (v16cbfloat16 a, v16cbfloat16 b, int sub_mul)
 

Multiplication of (2x8) with (8x2) bfloat16 and complex bfloat16 with dynamic negation of multiplication result, zeroing of acc1, and negation of acc1

v4caccfloat mac_2x8_8x2_conf (v16bfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v4caccfloat msc_2x8_8x2_conf (v16bfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v4caccfloat addmac_2x8_8x2_conf (v16bfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, v4caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v4caccfloat addmsc_2x8_8x2_conf (v16bfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, v4caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 

Multiplication of (2x8) with (8x2) complex bfloat16 and bfloat16 with dynamic negation of multiplication result, zeroing of acc1, and negation of acc1

v4caccfloat mac_2x8_8x2_conf (v16cbfloat16 v1, v16bfloat16 v2, v4caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v4caccfloat msc_2x8_8x2_conf (v16cbfloat16 v1, v16bfloat16 v2, v4caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v4caccfloat addmac_2x8_8x2_conf (v16cbfloat16 v1, v16bfloat16 v2, v4caccfloat acc1, v4caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v4caccfloat addmsc_2x8_8x2_conf (v16cbfloat16 v1, v16bfloat16 v2, v4caccfloat acc1, v4caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 

Multiplication of (2x8) with (8x2) complex bfloat16 and complex bfloat16 with dynamic negation of multiplication result, zeroing of acc1, and negation of acc1

v4caccfloat mac_2x8_8x2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v4caccfloat msc_2x8_8x2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v4caccfloat addmac_2x8_8x2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, v4caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v4caccfloat addmsc_2x8_8x2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v4caccfloat acc1, v4caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 

channel by channel multiplication of (1x2) with (2x1) complex bfloat16 and complex bfloat16

v8caccfloat mul_elem_8_2 (v16cbfloat16 v1, v16cbfloat16 v2)
 
v8caccfloat negmul_elem_8_2 (v16cbfloat16 v1, v16cbfloat16 v2)
 
v8caccfloat mac_elem_8_2 (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1)
 
v8caccfloat msc_elem_8_2 (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1)
 
v8caccfloat addmac_elem_8_2 (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2)
 
v8caccfloat addmsc_elem_8_2 (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2)
 

channel by channel multiplication of (1x2) with (2x1) bfloat16 and complex bfloat16

v8caccfloat mul_elem_8_2 (v16bfloat16 v1, v16cbfloat16 v2)
 
v8caccfloat negmul_elem_8_2 (v16bfloat16 v1, v16cbfloat16 v2)
 
v8caccfloat mac_elem_8_2 (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1)
 
v8caccfloat msc_elem_8_2 (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1)
 
v8caccfloat addmac_elem_8_2 (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2)
 
v8caccfloat addmsc_elem_8_2 (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2)
 

channel by channel multiplication of (1x2) with (2x1) complex bfloat16 and bfloat16

v8caccfloat mul_elem_8_2 (v16cbfloat16 v1, v16bfloat16 v2)
 
v8caccfloat negmul_elem_8_2 (v16cbfloat16 v1, v16bfloat16 v2)
 
v8caccfloat mac_elem_8_2 (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1)
 
v8caccfloat msc_elem_8_2 (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1)
 
v8caccfloat addmac_elem_8_2 (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, v8caccfloat acc2)
 
v8caccfloat addmsc_elem_8_2 (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, v8caccfloat acc2)
 

channel by channel multiplication of (1x2) with (2x1) for complex bfloat16 and complex bfloat16 with dynamic negation of multiplication result

v8caccfloat mul_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, int sub_mul)
 

channel by channel multiplication of (1x2) with (2x1) for complex bfloat16 and complex bfloat16 with dynamic negation of multiplication result, zeroing of acc1, and negation of acc1

v8caccfloat mac_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat submac_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat submsc_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 

channel by channel multiplication of (1x2) with (2x1) for complex cbfloat16 and bfloat16 with dynamic negation of multiplication result

v8caccfloat mul_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, int sub_mul)
 

channel by channel multiplication of (1x2) with (2x1) for complex bfloat16 and bfloat16 with dynamic negation of multiplication result, zeroing of acc1, and negation of acc1

v8caccfloat mac_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 

channel by channel multiplication of (1x2) with (2x1) for bfloat16 and complex bfloat16 with dynamic negation of multiplication result

v8caccfloat mul_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, int sub_mul)
 

channel by channel multiplication of (1x2) with (2x1) for bfloat16 and complex bfloat16 with dynamic negation of multiplication result, zeroing of acc1, and negation of acc1

v8caccfloat mac_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2)
 

Emulated vector float operations with dynamic negation of multiplication result, zeroing of acc1, negation of acc1

v8caccfloat mul_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, int sub_mask, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, int sub_mask, int sub_mul)
 
v8caccfloat mac_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16cbfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat mul_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, int sub_mask, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, int sub_mask, int sub_mul)
 
v8caccfloat mac_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16cbfloat16 v1, v16bfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat mul_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, int sub_mask, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, int sub_mask, int sub_mul)
 
v8caccfloat mac_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16bfloat16 v1, v16cbfloat16 v2, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 

Emulated vector float operations with dynamic sign, dynamic negation of multiplication result, zeroing of acc1, negation of acc1

v8caccfloat mul_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, int sub_mask, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, int sub_mask, int sub_mul)
 
v8caccfloat mac_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat mul_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16bfloat16 v2, int sgn_y, int sub_mask, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16bfloat16 v2, int sgn_y, int sub_mask, int sub_mul)
 
v8caccfloat mac_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16bfloat16 v2, int sgn_y, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16bfloat16 v2, int sgn_y, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16bfloat16 v2, int sgn_y, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16cbfloat16 v1, int sgn_x, v16bfloat16 v2, int sgn_y, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat mul_elem_8_2_conf (v16bfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, int sub_mask, int sub_mul)
 
v8caccfloat negmul_elem_8_2_conf (v16bfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, int sub_mask, int sub_mul)
 
v8caccfloat mac_elem_8_2_conf (v16bfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat msc_elem_8_2_conf (v16bfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1)
 
v8caccfloat addmac_elem_8_2_conf (v16bfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 
v8caccfloat addmsc_elem_8_2_conf (v16bfloat16 v1, int sgn_x, v16cbfloat16 v2, int sgn_y, v8caccfloat acc1, v8caccfloat acc2, int zero_acc1, int sub_mask, int sub_mul, int sub_acc1, int sub_acc2)
 

Function Documentation

◆ addmac_2x8_8x2() [1/3]

v4caccfloat addmac_2x8_8x2 ( v16bfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1,
v4caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmac_2x8_8x2() [2/3]

v4caccfloat addmac_2x8_8x2 ( v16cbfloat16  a,
v16bfloat16  b,
v4caccfloat  acc1,
v4caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmac_2x8_8x2() [3/3]

v4caccfloat addmac_2x8_8x2 ( v16cbfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1,
v4caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmac_2x8_8x2_conf() [1/3]

v4caccfloat addmac_2x8_8x2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
v4caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_2x8_8x2_conf() [2/3]

v4caccfloat addmac_2x8_8x2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v4caccfloat  acc1,
v4caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_2x8_8x2_conf() [3/3]

v4caccfloat addmac_2x8_8x2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
v4caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2() [1/3]

v8caccfloat addmac_elem_8_2 ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmac_elem_8_2() [2/3]

v8caccfloat addmac_elem_8_2 ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmac_elem_8_2() [3/3]

v8caccfloat addmac_elem_8_2 ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmac_elem_8_2_conf() [1/9]

v8caccfloat addmac_elem_8_2_conf ( v16bfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [2/9]

v8caccfloat addmac_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [3/9]

v8caccfloat addmac_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [4/9]

v8caccfloat addmac_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16bfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [5/9]

v8caccfloat addmac_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [6/9]

v8caccfloat addmac_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [7/9]

v8caccfloat addmac_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [8/9]

v8caccfloat addmac_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmac_elem_8_2_conf() [9/9]

v8caccfloat addmac_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_2x8_8x2() [1/3]

v4caccfloat addmsc_2x8_8x2 ( v16bfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1,
v4caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmsc_2x8_8x2() [2/3]

v4caccfloat addmsc_2x8_8x2 ( v16cbfloat16  a,
v16bfloat16  b,
v4caccfloat  acc1,
v4caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmsc_2x8_8x2() [3/3]

v4caccfloat addmsc_2x8_8x2 ( v16cbfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1,
v4caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmsc_2x8_8x2_conf() [1/3]

v4caccfloat addmsc_2x8_8x2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
v4caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_2x8_8x2_conf() [2/3]

v4caccfloat addmsc_2x8_8x2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v4caccfloat  acc1,
v4caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_2x8_8x2_conf() [3/3]

v4caccfloat addmsc_2x8_8x2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
v4caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2() [1/3]

v8caccfloat addmsc_elem_8_2 ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmsc_elem_8_2() [2/3]

v8caccfloat addmsc_elem_8_2 ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmsc_elem_8_2() [3/3]

v8caccfloat addmsc_elem_8_2 ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [1/9]

v8caccfloat addmsc_elem_8_2_conf ( v16bfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [2/9]

v8caccfloat addmsc_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [3/9]

v8caccfloat addmsc_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [4/9]

v8caccfloat addmsc_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16bfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [5/9]

v8caccfloat addmsc_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [6/9]

v8caccfloat addmsc_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [7/9]

v8caccfloat addmsc_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [8/9]

v8caccfloat addmsc_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
acc2Accumulator 2
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ addmsc_elem_8_2_conf() [9/9]

v8caccfloat addmsc_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ mac_2x8_8x2() [1/3]

v4caccfloat mac_2x8_8x2 ( v16bfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ mac_2x8_8x2() [2/3]

v4caccfloat mac_2x8_8x2 ( v16cbfloat16  a,
v16bfloat16  b,
v4caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ mac_2x8_8x2() [3/3]

v4caccfloat mac_2x8_8x2 ( v16cbfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ mac_2x8_8x2_conf() [1/3]

v4caccfloat mac_2x8_8x2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_2x8_8x2_conf() [2/3]

v4caccfloat mac_2x8_8x2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v4caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_2x8_8x2_conf() [3/3]

v4caccfloat mac_2x8_8x2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2() [1/3]

v8caccfloat mac_elem_8_2 ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ mac_elem_8_2() [2/3]

v8caccfloat mac_elem_8_2 ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ mac_elem_8_2() [3/3]

v8caccfloat mac_elem_8_2 ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ mac_elem_8_2_conf() [1/9]

v8caccfloat mac_elem_8_2_conf ( v16bfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2_conf() [2/9]

v8caccfloat mac_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2_conf() [3/9]

v8caccfloat mac_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mac_elem_8_2_conf() [4/9]

v8caccfloat mac_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16bfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2_conf() [5/9]

v8caccfloat mac_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2_conf() [6/9]

v8caccfloat mac_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2_conf() [7/9]

v8caccfloat mac_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mac_elem_8_2_conf() [8/9]

v8caccfloat mac_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ mac_elem_8_2_conf() [9/9]

v8caccfloat mac_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ msc_2x8_8x2() [1/3]

v4caccfloat msc_2x8_8x2 ( v16bfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ msc_2x8_8x2() [2/3]

v4caccfloat msc_2x8_8x2 ( v16cbfloat16  a,
v16bfloat16  b,
v4caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ msc_2x8_8x2() [3/3]

v4caccfloat msc_2x8_8x2 ( v16cbfloat16  a,
v16cbfloat16  b,
v4caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ msc_2x8_8x2_conf() [1/3]

v4caccfloat msc_2x8_8x2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_2x8_8x2_conf() [2/3]

v4caccfloat msc_2x8_8x2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v4caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_2x8_8x2_conf() [3/3]

v4caccfloat msc_2x8_8x2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v4caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2() [1/3]

v8caccfloat msc_elem_8_2 ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ msc_elem_8_2() [2/3]

v8caccfloat msc_elem_8_2 ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ msc_elem_8_2() [3/3]

v8caccfloat msc_elem_8_2 ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
Returns
Result of operation

◆ msc_elem_8_2_conf() [1/9]

v8caccfloat msc_elem_8_2_conf ( v16bfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2_conf() [2/9]

v8caccfloat msc_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2_conf() [3/9]

v8caccfloat msc_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ msc_elem_8_2_conf() [4/9]

v8caccfloat msc_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16bfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2_conf() [5/9]

v8caccfloat msc_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2_conf() [6/9]

v8caccfloat msc_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2_conf() [7/9]

v8caccfloat msc_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ msc_elem_8_2_conf() [8/9]

v8caccfloat msc_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mask,
int  sub_mul,
int  sub_acc1 
)
Parameters
v1Matrix A
v2Matrix B
acc1Accumulator 1
zero_acc1Zeroing mask for acc1
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
Returns
Result of operation

◆ msc_elem_8_2_conf() [9/9]

v8caccfloat msc_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
int  zero_acc1,
int  sub_mul,
int  sub_acc1 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_2x8_8x2() [1/3]

v4caccfloat mul_2x8_8x2 ( v16bfloat16  a,
v16cbfloat16  b 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ mul_2x8_8x2() [2/3]

v4caccfloat mul_2x8_8x2 ( v16cbfloat16  a,
v16bfloat16  b 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ mul_2x8_8x2() [3/3]

v4caccfloat mul_2x8_8x2 ( v16cbfloat16  a,
v16cbfloat16  b 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ mul_2x8_8x2_conf() [1/3]

v4caccfloat mul_2x8_8x2_conf ( v16bfloat16  a,
v16cbfloat16  b,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask for multiplication result
Returns
Result of operation

◆ mul_2x8_8x2_conf() [2/3]

v4caccfloat mul_2x8_8x2_conf ( v16cbfloat16  a,
v16bfloat16  b,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask for multiplication result
Returns
Result of operation

◆ mul_2x8_8x2_conf() [3/3]

v4caccfloat mul_2x8_8x2_conf ( v16cbfloat16  a,
v16cbfloat16  b,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask for multiplication result
Returns
Result of operation

◆ mul_elem_8_2() [1/3]

v8caccfloat mul_elem_8_2 ( v16bfloat16  v1,
v16cbfloat16  v2 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ mul_elem_8_2() [2/3]

v8caccfloat mul_elem_8_2 ( v16cbfloat16  v1,
v16bfloat16  v2 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ mul_elem_8_2() [3/3]

v8caccfloat mul_elem_8_2 ( v16cbfloat16  v1,
v16cbfloat16  v2 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ mul_elem_8_2_conf() [1/9]

v8caccfloat mul_elem_8_2_conf ( v16bfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [2/9]

v8caccfloat mul_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
v2Matrix B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [3/9]

v8caccfloat mul_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [4/9]

v8caccfloat mul_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16bfloat16  v2,
int  sgn_y,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [5/9]

v8caccfloat mul_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [6/9]

v8caccfloat mul_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
v2Matrix B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [7/9]

v8caccfloat mul_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [8/9]

v8caccfloat mul_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
v2Matrix B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ mul_elem_8_2_conf() [9/9]

v8caccfloat mul_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_2x8_8x2() [1/3]

v4caccfloat negmul_2x8_8x2 ( v16bfloat16  a,
v16cbfloat16  b 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ negmul_2x8_8x2() [2/3]

v4caccfloat negmul_2x8_8x2 ( v16cbfloat16  a,
v16bfloat16  b 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ negmul_2x8_8x2() [3/3]

v4caccfloat negmul_2x8_8x2 ( v16cbfloat16  a,
v16cbfloat16  b 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ negmul_2x8_8x2_conf() [1/3]

v4caccfloat negmul_2x8_8x2_conf ( v16bfloat16  a,
v16cbfloat16  b,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask for multiplication result
Returns
Result of operation

◆ negmul_2x8_8x2_conf() [2/3]

v4caccfloat negmul_2x8_8x2_conf ( v16cbfloat16  a,
v16bfloat16  b,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask for multiplication result
Returns
Result of operation

◆ negmul_2x8_8x2_conf() [3/3]

v4caccfloat negmul_2x8_8x2_conf ( v16cbfloat16  a,
v16cbfloat16  b,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask for multiplication result
Returns
Result of operation

◆ negmul_elem_8_2() [1/3]

v8caccfloat negmul_elem_8_2 ( v16bfloat16  v1,
v16cbfloat16  v2 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ negmul_elem_8_2() [2/3]

v8caccfloat negmul_elem_8_2 ( v16cbfloat16  v1,
v16bfloat16  v2 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ negmul_elem_8_2() [3/3]

v8caccfloat negmul_elem_8_2 ( v16cbfloat16  v1,
v16cbfloat16  v2 
)
Parameters
aMatrix A
bMatrix B
Returns
Result of operation

◆ negmul_elem_8_2_conf() [1/9]

v8caccfloat negmul_elem_8_2_conf ( v16bfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [2/9]

v8caccfloat negmul_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
v2Matrix B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [3/9]

v8caccfloat negmul_elem_8_2_conf ( v16bfloat16  v1,
v16cbfloat16  v2,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [4/9]

v8caccfloat negmul_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16bfloat16  v2,
int  sgn_y,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [5/9]

v8caccfloat negmul_elem_8_2_conf ( v16cbfloat16  v1,
int  sgn_x,
v16cbfloat16  v2,
int  sgn_y,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
sgn_xSign of input A
v2Matrix B
sgn_ySign of input B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [6/9]

v8caccfloat negmul_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
v2Matrix B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [7/9]

v8caccfloat negmul_elem_8_2_conf ( v16cbfloat16  v1,
v16bfloat16  v2,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [8/9]

v8caccfloat negmul_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
int  sub_mask,
int  sub_mul 
)
Parameters
v1Matrix A
v2Matrix B
sub_maskComplex negation mask of multiplication result
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ negmul_elem_8_2_conf() [9/9]

v8caccfloat negmul_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
int  sub_mul 
)
Parameters
aMatrix A
bMatrix B
sub_mulNegation mask of multiplication result
Returns
Result of operation

◆ submac_elem_8_2_conf()

v8caccfloat submac_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation

◆ submsc_elem_8_2_conf()

v8caccfloat submsc_elem_8_2_conf ( v16cbfloat16  v1,
v16cbfloat16  v2,
v8caccfloat  acc1,
v8caccfloat  acc2,
int  zero_acc1,
int  sub_mul,
int  sub_acc1,
int  sub_acc2 
)
Parameters
aMatrix A
bMatrix B
acc1Accumulator 1 input
acc2Accumulator 2 input
zero_acc1Zeroing mask for acc1
sub_mulNegation mask of multiplication result
sub_acc1Negation mask of acc1
sub_acc2Negation mask of acc2
Returns
Result of operation