![]() |
AI Engine-ML Intrinsics User Guide (v2024.1)
|
Matrix multiplications in which matrix A has data elements of 8 bit and matrix B has data elements of 4 bit. These operations are emulated on top of int8 x int8. More...
Matrix multiplications in which matrix A has data elements of 8 bit and matrix B has data elements of 4 bit. These operations are emulated on top of int8 x int8.
For an explanation how these operations works see Multiply Accumulate.
4x16_16x8 with dynamic negation of multiplication result <br> | |
v32acc32 | mul_4x16_16x8_conf (v64int8 a, v128int4 b, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64int8 a, v128int4 b, int sub_mul) |
v32acc32 | mul_4x16_16x8_conf (v64int8 a, v128uint4 b, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64int8 a, v128uint4 b, int sub_mul) |
v32acc32 | mul_4x16_16x8_conf (v64uint8 a, v128uint4 b, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64uint8 a, v128uint4 b, int sub_mul) |
v32acc32 | mul_4x16_16x8_conf (v64uint8 a, v128int4 b, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64uint8 a, v128int4 b, int sub_mul) |
4x16_16x8 with dynamic negation of multiplication result, zeroing of acc1, negation of acc1 | |
v32acc32 | mac_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | msc_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64int8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | mac_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | msc_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64int8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | mac_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | msc_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64uint8 a, v128uint4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | mac_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | msc_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64uint8 a, v128int4 b, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
4x16_16x8 with dynamic sign and dynamic negation of multiplication result | |
v32acc32 | mac_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | mul_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, int sub_mul) |
v32acc32 | msc_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64int8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | mac_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | mul_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, int sub_mul) |
v32acc32 | msc_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64int8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | mac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | mul_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, int sub_mul) |
v32acc32 | msc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128uint4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | mac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | mul_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, int sub_mul) |
v32acc32 | negmul_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, int sub_mul) |
v32acc32 | msc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int shift16, int sub_mul, int sub_acc1) |
v32acc32 | negmsc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | negmac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc, int zero_acc, int sub_mul, int sub_acc1) |
v32acc32 | addmac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | addmsc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int shift16, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submac_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 | submsc_4x16_16x8_conf (v64uint8 a, int sgn_x, v128int4 b, int sgn_y, v32acc32 acc1, v32acc32 acc2, int zero_acc1, int sub_mul, int sub_acc1, int sub_acc2) |
v32acc32 addmac_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmac_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmac_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmac_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmac_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmsc_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmsc_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmsc_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 addmsc_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
v32acc32 mac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 mac_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
v32acc32 msc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 msc_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | shift16, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
shift16 | Shift mask of input accumulator acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
v32acc32 negmac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmac_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
a | Matrix A |
b | Matrix B |
acc | Accumulator |
v32acc32 negmsc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 negmsc_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc, | ||
int | zero_acc, | ||
int | sub_mul, | ||
int | sub_acc1 | ||
) |
a | Matrix A |
b | Matrix B |
acc | Input accumulator |
zero_acc | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
b | Matrix B |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
sub_mul | Negation mask of multiplication result |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
a | Matrix A |
b | Matrix B |
sub_mul | Negation mask for multiplication result. If a bit of sub_mul is set the corresponding vector lane of the output accumulator will be negated. |
v32acc32 submac_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submac_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submac_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submac_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submac_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submsc_4x16_16x8 | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submsc_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submsc_4x16_16x8 | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
v32acc32 submsc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64int8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64int8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64int8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128int4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64uint8 | a, |
int | sgn_x, | ||
v128uint4 | b, | ||
int | sgn_y, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
sgn_x | Sign mask for matrix A |
b | Matrix B |
sgn_y | Sign mask for matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64uint8 | a, |
v128int4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |
v32acc32 submsc_4x16_16x8_conf | ( | v64uint8 | a, |
v128uint4 | b, | ||
v32acc32 | acc1, | ||
v32acc32 | acc2, | ||
int | zero_acc1, | ||
int | sub_mul, | ||
int | sub_acc1, | ||
int | sub_acc2 | ||
) |
a | Matrix A |
b | Matrix B |
acc1 | Input accumulator 1 |
acc2 | Input accumulator 2 |
zero_acc1 | Zeroing mask for acc1 |
sub_mul | Negation mask of multiplication result |
sub_acc1 | Negation mask of acc1 |