Motorola DSP96002 User Manual page 558

fmpy
fmpy.s d7,d3,d2
fmpy
fmpy
fmpy
fmpy.s d7,d3,d0
fmpy
fmpy
move
B.1.22 NxN NxN Matrix Multiply
The matrix multiplications are for square NxN matrices. All the elements are stored in "row major" format.
i.e. for the array A:
a(1,1) ... a(1,N)
the elements are stored:
a(1,1), a(1,2), ..., a(1,N), a(2,1), a(2,2), ..., a(2,N), ...
The following code implements C=AB where A and B are square matrices.
move #mat_a,r0
move #mat_b,r4
move #mat_c,r6
move #N,n0
move n0,n5
do #N,_rows
do #N,_cols
move r0,r1
move r4,r5
clr a
move
MOTOROLA
d7,d4,d0 fadd.s d2,d1
d7,d5,d2 fadd.s d2,d0
d7,d6,d2 fadd.s d2,d0
d7,d4,d1 fadd.s d2,d0
d7,d5,d0 fadd.s d0,d1
d7,d6,d0 fadd.s d0,d1
fadd.s d0,d1
a(N,1) ... a(N,N)
DSP56000 IMPLEMENTATION
x:(r1)+,x0 y:(r5)+n5,y0
DSP96002 USER'S MANUAL
y:(r4)+,d7.s ;b23
d1.s,x:(r1)+ y:(r4)+,d7.s ;b33
y:(r4)+,d7.s ;b43
y:(r4)+,d7.s ;b14
y:(r4)+,d7.s ;b24
d0.s,x:(r1)+ y:(r4)+,d7.s ;b34
y:(r4)+,d7.s ;b44
d1.s,x:(r1)+
;point to A
;point to B
;output mat C
;array size
;do rows
;do columns
;copy start of row A
;copy start of col B
;clear sum and pipe
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
--- ---
Totals: 19 19
Program ICycles
Words
1 1
1 1
1 1
1 1
1 1
2 3
2 3
1 1
1 1
1 1
1 1
B-39